├── Exploratory_Data_Analysis.ipynb ├── Images ├── 2in1.png ├── Accu.png ├── Accu2.png ├── Final dataset.PNG ├── Imbalance.png ├── Imbalance2.PNG ├── Jupyter_logo.png ├── Numpy_logo.png ├── Pandas_logo.png ├── Python_logo.png ├── RF.PNG ├── ROC.png ├── RP.png ├── Seaborn_logo.png ├── ab.PNG ├── abgb.PNG ├── contingency_table1.PNG ├── contingency_table2.PNG ├── contingency_table3.PNG ├── corr.png ├── corr1.png ├── corr11.PNG ├── corr111.png ├── corr2.PNG ├── corr22.png ├── corr222.png ├── corrplot.png ├── df.info.PNG ├── fet imp 2.PNG ├── fet imp.PNG ├── gb.PNG ├── knn.PNG ├── log reg.PNG ├── matplotlib_logo.png ├── null column.PNG ├── nulls.PNG ├── others.png ├── repeat.PNG ├── rfab.PNG └── svc.PNG ├── ML_backorders.ipynb ├── Preliminary_Data_Analysis.ipynb ├── README.md ├── Sample work report.pdf └── Wrangling_of_data.ipynb /Images/2in1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/2in1.png -------------------------------------------------------------------------------- /Images/Accu.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Accu.png -------------------------------------------------------------------------------- /Images/Accu2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Accu2.png -------------------------------------------------------------------------------- /Images/Final dataset.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Final dataset.PNG -------------------------------------------------------------------------------- /Images/Imbalance.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Imbalance.png -------------------------------------------------------------------------------- /Images/Imbalance2.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Imbalance2.PNG -------------------------------------------------------------------------------- /Images/Jupyter_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Jupyter_logo.png -------------------------------------------------------------------------------- /Images/Numpy_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Numpy_logo.png -------------------------------------------------------------------------------- /Images/Pandas_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Pandas_logo.png -------------------------------------------------------------------------------- /Images/Python_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Python_logo.png -------------------------------------------------------------------------------- /Images/RF.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/RF.PNG -------------------------------------------------------------------------------- /Images/ROC.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/ROC.png -------------------------------------------------------------------------------- /Images/RP.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/RP.png -------------------------------------------------------------------------------- /Images/Seaborn_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/Seaborn_logo.png -------------------------------------------------------------------------------- /Images/ab.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/ab.PNG -------------------------------------------------------------------------------- /Images/abgb.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/abgb.PNG -------------------------------------------------------------------------------- /Images/contingency_table1.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/contingency_table1.PNG -------------------------------------------------------------------------------- /Images/contingency_table2.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/contingency_table2.PNG -------------------------------------------------------------------------------- /Images/contingency_table3.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/contingency_table3.PNG -------------------------------------------------------------------------------- /Images/corr.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr.png -------------------------------------------------------------------------------- /Images/corr1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr1.png -------------------------------------------------------------------------------- /Images/corr11.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr11.PNG -------------------------------------------------------------------------------- /Images/corr111.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr111.png -------------------------------------------------------------------------------- /Images/corr2.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr2.PNG -------------------------------------------------------------------------------- /Images/corr22.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr22.png -------------------------------------------------------------------------------- /Images/corr222.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corr222.png -------------------------------------------------------------------------------- /Images/corrplot.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/corrplot.png -------------------------------------------------------------------------------- /Images/df.info.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/df.info.PNG -------------------------------------------------------------------------------- /Images/fet imp 2.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/fet imp 2.PNG -------------------------------------------------------------------------------- /Images/fet imp.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/fet imp.PNG -------------------------------------------------------------------------------- /Images/gb.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/gb.PNG -------------------------------------------------------------------------------- /Images/knn.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/knn.PNG -------------------------------------------------------------------------------- /Images/log reg.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/log reg.PNG -------------------------------------------------------------------------------- /Images/matplotlib_logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/matplotlib_logo.png -------------------------------------------------------------------------------- /Images/null column.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/null column.PNG -------------------------------------------------------------------------------- /Images/nulls.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/nulls.PNG -------------------------------------------------------------------------------- /Images/others.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/others.png -------------------------------------------------------------------------------- /Images/repeat.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/repeat.PNG -------------------------------------------------------------------------------- /Images/rfab.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/rfab.PNG -------------------------------------------------------------------------------- /Images/svc.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Images/svc.PNG -------------------------------------------------------------------------------- /Preliminary_Data_Analysis.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "# Preliminary Data Analysis with Sample Data" 8 | ] 9 | }, 10 | { 11 | "cell_type": "code", 12 | "execution_count": 48, 13 | "metadata": {}, 14 | "outputs": [], 15 | "source": [ 16 | "#Importing the packages\n", 17 | "import pandas as pd\n", 18 | "import numpy as np\n", 19 | "import matplotlib.pylab as plt\n", 20 | "import seaborn as sns\n", 21 | "from sklearn.model_selection import train_test_split\n", 22 | "from sklearn.linear_model import LogisticRegression\n", 23 | "from sklearn.metrics import classification_report\n", 24 | "from sklearn.metrics import accuracy_score\n", 25 | "from sklearn.metrics import roc_curve\n", 26 | "from sklearn.model_selection import cross_val_score\n", 27 | "from sklearn.model_selection import GridSearchCV\n", 28 | "from sklearn import svm\n", 29 | "from sklearn.model_selection import RandomizedSearchCV\n", 30 | "from sklearn.ensemble import RandomForestClassifier\n", 31 | "from sklearn.ensemble import GradientBoostingClassifier\n", 32 | "from sklearn.ensemble import AdaBoostClassifier\n", 33 | "from scipy import stats as st\n", 34 | "from sklearn.neighbors import KNeighborsClassifier\n", 35 | "from sklearn.preprocessing import StandardScaler\n", 36 | "from sklearn.pipeline import Pipeline\n", 37 | "from sklearn.metrics import precision_score\n", 38 | "from sklearn.metrics import average_precision_score\n", 39 | "from sklearn.metrics import precision_recall_curve\n", 40 | "%matplotlib inline" 41 | ] 42 | }, 43 | { 44 | "cell_type": "markdown", 45 | "metadata": {}, 46 | "source": [ 47 | "As a part of the preliminary data analysis, will be working on 30% of the data.\n", 48 | "\n", 49 | "1. Within the 10% of the data, will be splitting the data into train and test.\n", 50 | "\n", 51 | "2. Will be validating the models with the sample before working on the full datasets.\n", 52 | "\n", 53 | "3. Precision of the models are validated through the recall_precision_curve.\n" 54 | ] 55 | }, 56 | { 57 | "cell_type": "code", 58 | "execution_count": 51, 59 | "metadata": {}, 60 | "outputs": [], 61 | "source": [ 62 | "#Reading the clean file\n", 63 | "masterData = pd.read_csv('Backorder.csv')" 64 | ] 65 | }, 66 | { 67 | "cell_type": "code", 68 | "execution_count": 52, 69 | "metadata": {}, 70 | "outputs": [ 71 | { 72 | "data": { 73 | "image/png": "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\n", 74 | "text/plain": [ 75 | "
" 76 | ] 77 | }, 78 | "metadata": { 79 | "needs_background": "light" 80 | }, 81 | "output_type": "display_data" 82 | } 83 | ], 84 | "source": [ 85 | "plt.figure(figsize=(12,8))\n", 86 | "ax=sns.histplot(data=masterData,x='went_on_backorder',color='lightseagreen')\n", 87 | "ax.figure.savefig('Mean.png', dpi=500,bbox_inches='tight')" 88 | ] 89 | }, 90 | { 91 | "cell_type": "code", 92 | "execution_count": 53, 93 | "metadata": {}, 94 | "outputs": [ 95 | { 96 | "data": { 97 | "text/plain": [ 98 | "No 1628546\n", 99 | "Yes 11188\n", 100 | "Name: went_on_backorder, dtype: int64" 101 | ] 102 | }, 103 | "execution_count": 53, 104 | "metadata": {}, 105 | "output_type": "execute_result" 106 | } 107 | ], 108 | "source": [ 109 | "masterData['went_on_backorder'].value_counts()" 110 | ] 111 | }, 112 | { 113 | "cell_type": "code", 114 | "execution_count": 54, 115 | "metadata": {}, 116 | "outputs": [ 117 | { 118 | "data": { 119 | "text/html": [ 120 | "
\n", 121 | "\n", 134 | "\n", 135 | " \n", 136 | " \n", 137 | " \n", 138 | " \n", 139 | " \n", 140 | " \n", 141 | " \n", 142 | " \n", 143 | " \n", 144 | " \n", 145 | " \n", 146 | " \n", 147 | " \n", 148 | " \n", 149 | " \n", 150 | " \n", 151 | " \n", 152 | " \n", 153 | " \n", 154 | " \n", 155 | " \n", 156 | " \n", 157 | " \n", 158 | " \n", 159 | " \n", 160 | " \n", 161 | " \n", 162 | " \n", 163 | " \n", 164 | " \n", 165 | " \n", 166 | " \n", 167 | " \n", 168 | " \n", 169 | " \n", 170 | " \n", 171 | " \n", 172 | " \n", 173 | " \n", 174 | " \n", 175 | " \n", 176 | " \n", 177 | " \n", 178 | " \n", 179 | " \n", 180 | " \n", 181 | " \n", 182 | " \n", 183 | " \n", 184 | " \n", 185 | " \n", 186 | " \n", 187 | " \n", 188 | " \n", 189 | " \n", 190 | " \n", 191 | " \n", 192 | " \n", 193 | " \n", 194 | " \n", 195 | " \n", 196 | " \n", 197 | " \n", 198 | " \n", 199 | " \n", 200 | " \n", 201 | " \n", 202 | " \n", 203 | " \n", 204 | " \n", 205 | " \n", 206 | " \n", 207 | " \n", 208 | " \n", 209 | " \n", 210 | " \n", 211 | " \n", 212 | " \n", 213 | " \n", 214 | " \n", 215 | " \n", 216 | " \n", 217 | " \n", 218 | " \n", 219 | " \n", 220 | " \n", 221 | " \n", 222 | " \n", 223 | " \n", 224 | " \n", 225 | " \n", 226 | " \n", 227 | " \n", 228 | " \n", 229 | " \n", 230 | " \n", 231 | " \n", 232 | " \n", 233 | " \n", 234 | " \n", 235 | " \n", 236 | " \n", 237 | " \n", 238 | " \n", 239 | " \n", 240 | " \n", 241 | " \n", 242 | " \n", 243 | " \n", 244 | " \n", 245 | " \n", 246 | " \n", 247 | " \n", 248 | " \n", 249 | " \n", 250 | " \n", 251 | " \n", 252 | " \n", 253 | " \n", 254 | " \n", 255 | " \n", 256 | " \n", 257 | " \n", 258 | " \n", 259 | " \n", 260 | " \n", 261 | " \n", 262 | " \n", 263 | " \n", 264 | " \n", 265 | " \n", 266 | " \n", 267 | " \n", 268 | " \n", 269 | " \n", 270 | " \n", 271 | " \n", 272 | " \n", 273 | " \n", 274 | " \n", 275 | " \n", 276 | " \n", 277 | " \n", 278 | " \n", 279 | " \n", 280 | " \n", 281 | " \n", 282 | " \n", 283 | " \n", 284 | " \n", 285 | " \n", 286 | " \n", 287 | " \n", 288 | " \n", 289 | " \n", 290 | " \n", 291 | " \n", 292 | " \n", 293 | " \n", 294 | " \n", 295 | " \n", 296 | " \n", 297 | " \n", 298 | " \n", 299 | " \n", 300 | " \n", 301 | " \n", 302 | " \n", 303 | " \n", 304 | " \n", 305 | " \n", 306 | " \n", 307 | " \n", 308 | " \n", 309 | " \n", 310 | " \n", 311 | " \n", 312 | " \n", 313 | " \n", 314 | " \n", 315 | " \n", 316 | " \n", 317 | " \n", 318 | " \n", 319 | " \n", 320 | " \n", 321 | " \n", 322 | " \n", 323 | " \n", 324 | " \n", 325 | " \n", 326 | " \n", 327 | " \n", 328 | " \n", 329 | " \n", 330 | " \n", 331 | " \n", 332 | " \n", 333 | " \n", 334 | " \n", 335 | " \n", 336 | " \n", 337 | " \n", 338 | " \n", 339 | " \n", 340 | " \n", 341 | " \n", 342 | " \n", 343 | " \n", 344 | " \n", 345 | " \n", 346 | " \n", 347 | " \n", 348 | " \n", 349 | " \n", 350 | " \n", 351 | " \n", 352 | " \n", 353 | " \n", 354 | " \n", 355 | " \n", 356 | " \n", 357 | " \n", 358 | " \n", 359 | " \n", 360 | " \n", 361 | " \n", 362 | " \n", 363 | " \n", 364 | " \n", 365 | " \n", 366 | " \n", 367 | " \n", 368 | " \n", 369 | " \n", 370 | " \n", 371 | " \n", 372 | " \n", 373 | " \n", 374 | " \n", 375 | " \n", 376 | " \n", 377 | " \n", 378 | " \n", 379 | "
product_idcurrent_inventorytransit_durationprior_sales_1_monthprior_sales_3_monthprior_sales_6_monthprior_sales_9_monthminimum_recommend_stocksource_has_issuesource_performance_6_monthssource_performance_12_monthsdeck_riskoe_constraintppap_riskstop_auto_buyrev_stopwent_on_backorder
022318490.02.00.04.012.023.02.0No0.000.00NoNoNoYesNoYes
1188455613.08.07.016.026.035.02.0No0.670.39NoNoNoYesNoYes
218273950.02.00.00.00.00.00.0No0.000.00YesNoYesYesNoYes
332755153.08.03.08.015.023.04.0No0.830.85NoNoNoYesNoNo
419889450.02.00.00.01.01.00.0No0.830.86YesNoNoNoNoYes
......................................................
2218322974620.012.00.02.04.010.00.0No0.730.76NoNoNoYesNoYes
2218418302210.08.00.00.00.00.00.0No0.980.98YesNoYesYesNoYes
2218520889750.09.02.03.06.07.00.0No0.700.66NoNoNoYesNoYes
22186112982916.09.00.07.07.08.00.0No0.690.68NoNoYesYesNoNo
221873104959-11.02.0343.0664.0779.0814.089.0No0.730.70NoNoNoYesNoYes
\n", 380 | "

22188 rows × 17 columns

\n", 381 | "
" 382 | ], 383 | "text/plain": [ 384 | " product_id current_inventory transit_duration prior_sales_1_month \\\n", 385 | "0 2231849 0.0 2.0 0.0 \n", 386 | "1 1884556 13.0 8.0 7.0 \n", 387 | "2 1827395 0.0 2.0 0.0 \n", 388 | "3 3275515 3.0 8.0 3.0 \n", 389 | "4 1988945 0.0 2.0 0.0 \n", 390 | "... ... ... ... ... \n", 391 | "22183 2297462 0.0 12.0 0.0 \n", 392 | "22184 1830221 0.0 8.0 0.0 \n", 393 | "22185 2088975 0.0 9.0 2.0 \n", 394 | "22186 1129829 16.0 9.0 0.0 \n", 395 | "22187 3104959 -11.0 2.0 343.0 \n", 396 | "\n", 397 | " prior_sales_3_month prior_sales_6_month prior_sales_9_month \\\n", 398 | "0 4.0 12.0 23.0 \n", 399 | "1 16.0 26.0 35.0 \n", 400 | "2 0.0 0.0 0.0 \n", 401 | "3 8.0 15.0 23.0 \n", 402 | "4 0.0 1.0 1.0 \n", 403 | "... ... ... ... \n", 404 | "22183 2.0 4.0 10.0 \n", 405 | "22184 0.0 0.0 0.0 \n", 406 | "22185 3.0 6.0 7.0 \n", 407 | "22186 7.0 7.0 8.0 \n", 408 | "22187 664.0 779.0 814.0 \n", 409 | "\n", 410 | " minimum_recommend_stock source_has_issue source_performance_6_months \\\n", 411 | "0 2.0 No 0.00 \n", 412 | "1 2.0 No 0.67 \n", 413 | "2 0.0 No 0.00 \n", 414 | "3 4.0 No 0.83 \n", 415 | "4 0.0 No 0.83 \n", 416 | "... ... ... ... \n", 417 | "22183 0.0 No 0.73 \n", 418 | "22184 0.0 No 0.98 \n", 419 | "22185 0.0 No 0.70 \n", 420 | "22186 0.0 No 0.69 \n", 421 | "22187 89.0 No 0.73 \n", 422 | "\n", 423 | " source_performance_12_months deck_risk oe_constraint ppap_risk \\\n", 424 | "0 0.00 No No No \n", 425 | "1 0.39 No No No \n", 426 | "2 0.00 Yes No Yes \n", 427 | "3 0.85 No No No \n", 428 | "4 0.86 Yes No No \n", 429 | "... ... ... ... ... \n", 430 | "22183 0.76 No No No \n", 431 | "22184 0.98 Yes No Yes \n", 432 | "22185 0.66 No No No \n", 433 | "22186 0.68 No No Yes \n", 434 | "22187 0.70 No No No \n", 435 | "\n", 436 | " stop_auto_buy rev_stop went_on_backorder \n", 437 | "0 Yes No Yes \n", 438 | "1 Yes No Yes \n", 439 | "2 Yes No Yes \n", 440 | "3 Yes No No \n", 441 | "4 No No Yes \n", 442 | "... ... ... ... \n", 443 | "22183 Yes No Yes \n", 444 | "22184 Yes No Yes \n", 445 | "22185 Yes No Yes \n", 446 | "22186 Yes No No \n", 447 | "22187 Yes No Yes \n", 448 | "\n", 449 | "[22188 rows x 17 columns]" 450 | ] 451 | }, 452 | "execution_count": 54, 453 | "metadata": {}, 454 | "output_type": "execute_result" 455 | } 456 | ], 457 | "source": [ 458 | "join1=masterData[masterData['went_on_backorder']=='Yes']\n", 459 | "join2=masterData[masterData['went_on_backorder']=='No']\n", 460 | "join3=join2.sample(n=11000, random_state = 2)\n", 461 | "\n", 462 | "masterData=pd.merge(join1,join3,how='outer')\n", 463 | "masterData=masterData.sample(frac=1).reset_index(drop=True)\n", 464 | "masterData" 465 | ] 466 | }, 467 | { 468 | "cell_type": "code", 469 | "execution_count": 55, 470 | "metadata": {}, 471 | "outputs": [ 472 | { 473 | "data": { 474 | "text/plain": [ 475 | "((22188, 17), (22188,))" 476 | ] 477 | }, 478 | "execution_count": 55, 479 | "metadata": {}, 480 | "output_type": "execute_result" 481 | } 482 | ], 483 | "source": [ 484 | "#Preparing the data for modeling\n", 485 | "#X = pd.DataFrame(masterData.columns)\n", 486 | "X = masterData[masterData.columns]\n", 487 | "y = masterData['went_on_backorder']\n", 488 | "X.shape, y.shape" 489 | ] 490 | }, 491 | { 492 | "cell_type": "code", 493 | "execution_count": 56, 494 | "metadata": {}, 495 | "outputs": [ 496 | { 497 | "data": { 498 | "text/plain": [ 499 | "((22188, 10), (22188,))" 500 | ] 501 | }, 502 | "execution_count": 56, 503 | "metadata": {}, 504 | "output_type": "execute_result" 505 | } 506 | ], 507 | "source": [ 508 | "#Convert the categorical data into numerical\n", 509 | "y = y.replace('Yes', 1)\n", 510 | "y = y.replace('No', 0)\n", 511 | "X['source_has_issue'] = X.source_has_issue.map({'Yes':1, 'No':0})\n", 512 | "X['deck_risk'] = X.deck_risk.map({'Yes':1, 'No':0})\n", 513 | "X['oe_constraint'] = X.oe_constraint.map({'Yes':1, 'No':0})\n", 514 | "X['ppap_risk'] = X.ppap_risk.map({'Yes':1, 'No':0})\n", 515 | "X['stop_auto_buy'] = X.stop_auto_buy.map({'Yes':1, 'No':0})\n", 516 | "X['rev_stop'] = X.rev_stop.map({'Yes':1, 'No':0})\n", 517 | "X = X.drop(['went_on_backorder', 'source_has_issue', 'rev_stop', 'deck_risk', 'ppap_risk', 'stop_auto_buy','product_id'], axis = 1, errors = 'ignore')\n", 518 | "X.shape, y.shape" 519 | ] 520 | }, 521 | { 522 | "cell_type": "code", 523 | "execution_count": 57, 524 | "metadata": {}, 525 | "outputs": [ 526 | { 527 | "data": { 528 | "text/plain": [ 529 | "((16641, 10), (16641,), (5547, 10), (5547,))" 530 | ] 531 | }, 532 | "execution_count": 57, 533 | "metadata": {}, 534 | "output_type": "execute_result" 535 | } 536 | ], 537 | "source": [ 538 | "#get the 10% of the data\n", 539 | "X_sample, X_data, y_sample, y_data = train_test_split(X, y, test_size = 0.25, random_state = 42)\n", 540 | "X_sample.shape, y_sample.shape, X_data.shape, y_data.shape" 541 | ] 542 | }, 543 | { 544 | "cell_type": "code", 545 | "execution_count": 58, 546 | "metadata": {}, 547 | "outputs": [ 548 | { 549 | "data": { 550 | "text/plain": [ 551 | "((11648, 10), (4993, 10), (11648,), (4993,))" 552 | ] 553 | }, 554 | "execution_count": 58, 555 | "metadata": {}, 556 | "output_type": "execute_result" 557 | } 558 | ], 559 | "source": [ 560 | "X_train, X_test, y_train, y_test = train_test_split(X_sample, y_sample, test_size = 0.30, random_state = 42)\n", 561 | "X_train.shape, X_test.shape, y_train.shape, y_test.shape" 562 | ] 563 | }, 564 | { 565 | "cell_type": "markdown", 566 | "metadata": {}, 567 | "source": [ 568 | "# Training the KNN Model Classifier" 569 | ] 570 | }, 571 | { 572 | "cell_type": "code", 573 | "execution_count": 59, 574 | "metadata": {}, 575 | "outputs": [ 576 | { 577 | "name": "stdout", 578 | "output_type": "stream", 579 | "text": [ 580 | " precision recall f1-score support\n", 581 | "\n", 582 | " 0 0.75 0.70 0.72 2484\n", 583 | " 1 0.72 0.77 0.75 2509\n", 584 | "\n", 585 | " accuracy 0.74 4993\n", 586 | " macro avg 0.74 0.73 0.73 4993\n", 587 | "weighted avg 0.74 0.74 0.73 4993\n", 588 | "\n" 589 | ] 590 | } 591 | ], 592 | "source": [ 593 | "steps = [('scaler', StandardScaler()),\n", 594 | " ('knn', KNeighborsClassifier(n_jobs = -1))]\n", 595 | "#Create the pipeline: pipeline\n", 596 | "pipeline_knn = Pipeline(steps)\n", 597 | "\n", 598 | "# Fit to the training set\n", 599 | "pipeline_knn.fit(X_train, y_train)\n", 600 | "knn_prediction = pipeline_knn.predict( X_test )\n", 601 | "# get the classification report\n", 602 | "knn_report = classification_report( y_test, knn_prediction )\n", 603 | "#print the report\n", 604 | "print(knn_report)" 605 | ] 606 | }, 607 | { 608 | "cell_type": "markdown", 609 | "metadata": {}, 610 | "source": [ 611 | "# Training the SVC Model Classifier" 612 | ] 613 | }, 614 | { 615 | "cell_type": "code", 616 | "execution_count": 60, 617 | "metadata": {}, 618 | "outputs": [ 619 | { 620 | "name": "stdout", 621 | "output_type": "stream", 622 | "text": [ 623 | " precision recall f1-score support\n", 624 | "\n", 625 | " 0 0.59 0.70 0.64 2484\n", 626 | " 1 0.63 0.51 0.57 2509\n", 627 | "\n", 628 | " accuracy 0.61 4993\n", 629 | " macro avg 0.61 0.61 0.60 4993\n", 630 | "weighted avg 0.61 0.61 0.60 4993\n", 631 | "\n" 632 | ] 633 | } 634 | ], 635 | "source": [ 636 | "steps = [('scaler', StandardScaler()),\n", 637 | " ('svc', svm.SVC(class_weight = 'balanced'))]\n", 638 | "#Create the pipeline: pipeline\n", 639 | "pipeline_svc = Pipeline(steps)\n", 640 | "\n", 641 | "# Fit to the training set\n", 642 | "pipeline_svc.fit(X_train, y_train)\n", 643 | "svc_prediction = pipeline_svc.predict( X_test )\n", 644 | "# get the classification report\n", 645 | "svc_report = classification_report( y_test, svc_prediction )\n", 646 | "#print the report\n", 647 | "print(svc_report)" 648 | ] 649 | }, 650 | { 651 | "cell_type": "markdown", 652 | "metadata": {}, 653 | "source": [ 654 | "# Training the LogisticRegression Model " 655 | ] 656 | }, 657 | { 658 | "cell_type": "code", 659 | "execution_count": 61, 660 | "metadata": {}, 661 | "outputs": [ 662 | { 663 | "name": "stdout", 664 | "output_type": "stream", 665 | "text": [ 666 | " precision recall f1-score support\n", 667 | "\n", 668 | " 0 0.60 0.65 0.62 2484\n", 669 | " 1 0.62 0.58 0.60 2509\n", 670 | "\n", 671 | " accuracy 0.61 4993\n", 672 | " macro avg 0.61 0.61 0.61 4993\n", 673 | "weighted avg 0.61 0.61 0.61 4993\n", 674 | "\n" 675 | ] 676 | } 677 | ], 678 | "source": [ 679 | "steps = [('scaler', StandardScaler()),\n", 680 | " ('logreg', LogisticRegression(class_weight = 'balanced'))]\n", 681 | "#Create the pipeline: pipeline\n", 682 | "pipeline_logreg = Pipeline(steps)\n", 683 | "\n", 684 | "# Fit to the training set\n", 685 | "pipeline_logreg.fit(X_train, y_train)\n", 686 | "logreg_prediction = pipeline_logreg.predict( X_test )\n", 687 | "# get the classification report\n", 688 | "logreg_report = classification_report( y_test, logreg_prediction )\n", 689 | "#print the report\n", 690 | "print(logreg_report)" 691 | ] 692 | }, 693 | { 694 | "cell_type": "markdown", 695 | "metadata": {}, 696 | "source": [ 697 | "# Training the RandomForest Classifier" 698 | ] 699 | }, 700 | { 701 | "cell_type": "code", 702 | "execution_count": 62, 703 | "metadata": {}, 704 | "outputs": [ 705 | { 706 | "name": "stdout", 707 | "output_type": "stream", 708 | "text": [ 709 | " precision recall f1-score support\n", 710 | "\n", 711 | " 0 0.91 0.86 0.89 2484\n", 712 | " 1 0.87 0.92 0.89 2509\n", 713 | "\n", 714 | " accuracy 0.89 4993\n", 715 | " macro avg 0.89 0.89 0.89 4993\n", 716 | "weighted avg 0.89 0.89 0.89 4993\n", 717 | "\n" 718 | ] 719 | } 720 | ], 721 | "source": [ 722 | "#Train default RandomForest on 30% of the data\n", 723 | "rfmodel = RandomForestClassifier(random_state = 42)\n", 724 | "#Fit the training set\n", 725 | "rfmodel.fit(X_train, y_train)\n", 726 | "rfmodel_prediction = rfmodel.predict(X_test)\n", 727 | "#get the classification report\n", 728 | "rfmodel_report = classification_report(y_test, rfmodel_prediction)\n", 729 | "#print the report\n", 730 | "print(rfmodel_report)" 731 | ] 732 | }, 733 | { 734 | "cell_type": "markdown", 735 | "metadata": {}, 736 | "source": [ 737 | "# Training GradientBoosting Classifier model" 738 | ] 739 | }, 740 | { 741 | "cell_type": "code", 742 | "execution_count": 63, 743 | "metadata": {}, 744 | "outputs": [ 745 | { 746 | "name": "stdout", 747 | "output_type": "stream", 748 | "text": [ 749 | " precision recall f1-score support\n", 750 | "\n", 751 | " 0 0.88 0.85 0.86 2484\n", 752 | " 1 0.85 0.88 0.87 2509\n", 753 | "\n", 754 | " accuracy 0.87 4993\n", 755 | " macro avg 0.87 0.87 0.87 4993\n", 756 | "weighted avg 0.87 0.87 0.87 4993\n", 757 | "\n" 758 | ] 759 | } 760 | ], 761 | "source": [ 762 | "#Train default RandomForest on 30% of the data\n", 763 | "gbmodel = GradientBoostingClassifier(random_state = 42)\n", 764 | "#Fit the training set\n", 765 | "gbmodel.fit(X_train, y_train)\n", 766 | "gbmodel_prediction = gbmodel.predict(X_test)\n", 767 | "#get the classification report\n", 768 | "gbmodel_report = classification_report(y_test, gbmodel_prediction)\n", 769 | "#print the report\n", 770 | "print(gbmodel_report)" 771 | ] 772 | }, 773 | { 774 | "cell_type": "markdown", 775 | "metadata": {}, 776 | "source": [ 777 | "# Training AdaBoostClassifier Model" 778 | ] 779 | }, 780 | { 781 | "cell_type": "code", 782 | "execution_count": 64, 783 | "metadata": {}, 784 | "outputs": [ 785 | { 786 | "name": "stdout", 787 | "output_type": "stream", 788 | "text": [ 789 | " precision recall f1-score support\n", 790 | "\n", 791 | " 0 0.87 0.82 0.84 2484\n", 792 | " 1 0.83 0.87 0.85 2509\n", 793 | "\n", 794 | " accuracy 0.85 4993\n", 795 | " macro avg 0.85 0.85 0.85 4993\n", 796 | "weighted avg 0.85 0.85 0.85 4993\n", 797 | "\n" 798 | ] 799 | } 800 | ], 801 | "source": [ 802 | "#Train default RandomForest on 30% of the data\n", 803 | "abmodel = AdaBoostClassifier(random_state = 42)\n", 804 | "#Fit the training set\n", 805 | "abmodel.fit(X_train, y_train)\n", 806 | "abmodel_prediction = abmodel.predict(X_test)\n", 807 | "#get the classification report\n", 808 | "abmodel_report = classification_report(y_test, abmodel_prediction)\n", 809 | "#print the report\n", 810 | "print(abmodel_report)" 811 | ] 812 | }, 813 | { 814 | "cell_type": "markdown", 815 | "metadata": {}, 816 | "source": [ 817 | "# Training the AdaBoostClassifier with RandomForest Estimator" 818 | ] 819 | }, 820 | { 821 | "cell_type": "code", 822 | "execution_count": 65, 823 | "metadata": {}, 824 | "outputs": [ 825 | { 826 | "name": "stdout", 827 | "output_type": "stream", 828 | "text": [ 829 | " precision recall f1-score support\n", 830 | "\n", 831 | " 0 0.92 0.86 0.89 2484\n", 832 | " 1 0.87 0.92 0.90 2509\n", 833 | "\n", 834 | " accuracy 0.89 4993\n", 835 | " macro avg 0.90 0.89 0.89 4993\n", 836 | "weighted avg 0.90 0.89 0.89 4993\n", 837 | "\n" 838 | ] 839 | } 840 | ], 841 | "source": [ 842 | "#Train default RandomForest on 30% of the data\n", 843 | "abmodel_rf = AdaBoostClassifier(base_estimator = rfmodel, random_state = 42)\n", 844 | "#Fit the training set\n", 845 | "abmodel_rf.fit(X_train, y_train)\n", 846 | "abmodel_prediction_rf = abmodel_rf.predict(X_test)\n", 847 | "#get the classification report\n", 848 | "abmodel_report_rf = classification_report(y_test, abmodel_prediction_rf)\n", 849 | "#print the report\n", 850 | "print(abmodel_report_rf)" 851 | ] 852 | }, 853 | { 854 | "cell_type": "code", 855 | "execution_count": 43, 856 | "metadata": {}, 857 | "outputs": [ 858 | { 859 | "data": { 860 | "text/plain": [ 861 | "current_inventory 0.409697\n", 862 | "prior_sales_6_month 0.091196\n", 863 | "source_performance_6_months 0.086669\n", 864 | "source_performance_12_months 0.086269\n", 865 | "prior_sales_3_month 0.084116\n", 866 | "prior_sales_1_month 0.073360\n", 867 | "prior_sales_9_month 0.072090\n", 868 | "minimum_recommend_stock 0.049787\n", 869 | "transit_duration 0.046733\n", 870 | "oe_constraint 0.000082\n", 871 | "dtype: float64" 872 | ] 873 | }, 874 | "execution_count": 43, 875 | "metadata": {}, 876 | "output_type": "execute_result" 877 | } 878 | ], 879 | "source": [ 880 | "feature_import_rf = pd.Series(rfmodel.feature_importances_, index = X_test.columns)\n", 881 | "feature_import_rf.sort_values(ascending = False)\n" 882 | ] 883 | }, 884 | { 885 | "cell_type": "code", 886 | "execution_count": 44, 887 | "metadata": {}, 888 | "outputs": [ 889 | { 890 | "data": { 891 | "text/plain": [ 892 | "current_inventory 0.628698\n", 893 | "prior_sales_1_month 0.125767\n", 894 | "prior_sales_3_month 0.087195\n", 895 | "prior_sales_6_month 0.061263\n", 896 | "prior_sales_9_month 0.038240\n", 897 | "transit_duration 0.026493\n", 898 | "source_performance_6_months 0.015331\n", 899 | "source_performance_12_months 0.009316\n", 900 | "minimum_recommend_stock 0.007697\n", 901 | "oe_constraint 0.000000\n", 902 | "dtype: float64" 903 | ] 904 | }, 905 | "execution_count": 44, 906 | "metadata": {}, 907 | "output_type": "execute_result" 908 | } 909 | ], 910 | "source": [ 911 | "feature_import_gb = pd.Series(gbmodel.feature_importances_, index = X_test.columns)\n", 912 | "feature_import_gb.sort_values(ascending = False)\n" 913 | ] 914 | }, 915 | { 916 | "cell_type": "markdown", 917 | "metadata": {}, 918 | "source": [ 919 | "# Traning the AdaBoostClassifier with GradientBoosting Estimator" 920 | ] 921 | }, 922 | { 923 | "cell_type": "code", 924 | "execution_count": 45, 925 | "metadata": {}, 926 | "outputs": [ 927 | { 928 | "name": "stdout", 929 | "output_type": "stream", 930 | "text": [ 931 | " precision recall f1-score support\n", 932 | "\n", 933 | " 0 0.90 0.85 0.87 2472\n", 934 | " 1 0.86 0.90 0.88 2521\n", 935 | "\n", 936 | " accuracy 0.88 4993\n", 937 | " macro avg 0.88 0.88 0.88 4993\n", 938 | "weighted avg 0.88 0.88 0.88 4993\n", 939 | "\n" 940 | ] 941 | } 942 | ], 943 | "source": [ 944 | "#Train default AdaBoost model with the GB estimator\n", 945 | "abmodel_gb = AdaBoostClassifier(base_estimator = gbmodel, random_state = 42)\n", 946 | "#Fit the training model\n", 947 | "abmodel_gb.fit(X_train, y_train)\n", 948 | "abmodel_prediction_gb = abmodel_gb.predict(X_test)\n", 949 | "#get the classication report\n", 950 | "abmodel_report_gb = classification_report(y_test, abmodel_prediction_gb)\n", 951 | "#print the report\n", 952 | "print(abmodel_report_gb)" 953 | ] 954 | }, 955 | { 956 | "cell_type": "code", 957 | "execution_count": 46, 958 | "metadata": {}, 959 | "outputs": [ 960 | { 961 | "data": { 962 | "text/plain": [ 963 | "Text(0.5, 1.0, 'ROC Curve')" 964 | ] 965 | }, 966 | "execution_count": 46, 967 | "metadata": {}, 968 | "output_type": "execute_result" 969 | }, 970 | { 971 | "data": { 972 | "image/png": "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\n", 973 | "text/plain": [ 974 | "
" 975 | ] 976 | }, 977 | "metadata": { 978 | "needs_background": "light" 979 | }, 980 | "output_type": "display_data" 981 | } 982 | ], 983 | "source": [ 984 | "#plot ROC Curves\n", 985 | "fig , ax1 = plt.subplots(figsize=(9,9) )\n", 986 | "plt.plot([0, 1], [0, 1], 'k--')\n", 987 | "estimators =[pipeline_knn, pipeline_logreg, rfmodel, gbmodel, abmodel ]\n", 988 | "classifiers=['KNN', 'LogisticRegression', 'RandomForestClassifier','GradientBoostingClassifier',\n", 989 | " 'AdaBoostClassifier']\n", 990 | "colors = ['b', 'g', 'r', 'c', 'm']\n", 991 | "for i, estimator in enumerate(estimators):\n", 992 | " y_pred_prob = estimator.predict_proba(X_test)[:,1]\n", 993 | " fpr, tpr, thresholds= roc_curve(y_test, y_pred_prob)\n", 994 | " plt.plot(fpr, tpr, label=classifiers[i],color=colors[i]) \n", 995 | "plt.xlabel('False Positive Rate')\n", 996 | "plt.ylabel('True Positive Rate')\n", 997 | "plt.legend(loc=4)\n", 998 | "plt.title('ROC Curve')" 999 | ] 1000 | }, 1001 | { 1002 | "cell_type": "code", 1003 | "execution_count": 47, 1004 | "metadata": {}, 1005 | "outputs": [ 1006 | { 1007 | "data": { 1008 | "text/plain": [ 1009 | "Text(0.5, 1.0, 'Recall Precision Curve')" 1010 | ] 1011 | }, 1012 | "execution_count": 47, 1013 | "metadata": {}, 1014 | "output_type": "execute_result" 1015 | }, 1016 | { 1017 | "data": { 1018 | "image/png": "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\n", 1019 | "text/plain": [ 1020 | "
" 1021 | ] 1022 | }, 1023 | "metadata": { 1024 | "needs_background": "light" 1025 | }, 1026 | "output_type": "display_data" 1027 | } 1028 | ], 1029 | "source": [ 1030 | "#plot Recall-Precision Curves\n", 1031 | "fig , ax1 = plt.subplots(figsize=(9,9) )\n", 1032 | "estimators =[pipeline_knn, pipeline_logreg, rfmodel, gbmodel, abmodel ]\n", 1033 | "classifiers=['KNN', 'LogisticRegression', 'RandomForestClassifier','GradientBoostingClassifier',\n", 1034 | " 'AdaBoostClassifier']\n", 1035 | "colors = ['b', 'g', 'r', 'c', 'm']\n", 1036 | "for i, estimator in enumerate(estimators):\n", 1037 | " y_pred_prob = estimator.predict_proba(X_test)[:,1]\n", 1038 | " precision, recall, _ = precision_recall_curve(y_test,y_pred_prob)\n", 1039 | " average_precision= average_precision_score(y_test, y_pred_prob, average=\"micro\")\n", 1040 | " plt.plot(recall, precision, label='%s (average=%.3f)'%(classifiers[i],average_precision), color=colors[i])\n", 1041 | "plt.xlabel('Recall')\n", 1042 | "plt.ylabel('Precision')\n", 1043 | "plt.legend(loc=1)\n", 1044 | "plt.title('Recall Precision Curve')" 1045 | ] 1046 | }, 1047 | { 1048 | "cell_type": "code", 1049 | "execution_count": null, 1050 | "metadata": {}, 1051 | "outputs": [], 1052 | "source": [] 1053 | } 1054 | ], 1055 | "metadata": { 1056 | "kernelspec": { 1057 | "display_name": "Python 3", 1058 | "language": "python", 1059 | "name": "python3" 1060 | }, 1061 | "language_info": { 1062 | "codemirror_mode": { 1063 | "name": "ipython", 1064 | "version": 3 1065 | }, 1066 | "file_extension": ".py", 1067 | "mimetype": "text/x-python", 1068 | "name": "python", 1069 | "nbconvert_exporter": "python", 1070 | "pygments_lexer": "ipython3", 1071 | "version": "3.8.8" 1072 | } 1073 | }, 1074 | "nbformat": 4, 1075 | "nbformat_minor": 4 1076 | } 1077 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Backorders_Supply_Chain_Analysis (Competition Dataset) 2 | Analyzed the supply chain process-data and developed ML models using Python that can predict if a given order goes to back order with 89% precision and 46% recall in Random Forest model. 3 | 4 | ![](Images/corr.png) 5 | -------------------------------------------------------------------------------- /Sample work report.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/prs98/Backorders_Supply_Chain_Analysis/45489dd0c3887d4a0ba4e8b7e7d1cfd140898553/Sample work report.pdf --------------------------------------------------------------------------------