├── README.md
└── Stock_price_predection.ipynb
/README.md:
--------------------------------------------------------------------------------
1 | # 📈 Google Stock Price Prediction by using LSTM
2 |
3 | This project uses **Long Short-Term Memory (LSTM)**, a type of Recurrent Neural Network (RNN), to predict future stock prices of **Google (GOOGLE)** based on historical data. LSTM models are effective for time series forecasting due to their ability to learn long-term dependencies
4 |
5 | ---
6 | ## Author : Shashank Pandey
7 |
8 |
9 | ## 📊 Project Overview
10 |
11 | Stock price prediction is a challenging and fascinating task. In this project, we use an LSTM-based deep learning model to predict Google's stock prices using its historical data.
12 |
13 | The model is trained on past stock prices and aims to predict the stock's future closing prices. It provides a basic but powerful demonstration of how deep learning can be used in financial forecasting.
14 |
15 | ---
16 |
17 | ## 🧾 Dataset.
18 |
19 | - Source: Yahoo Finance (GOOGL stock).
20 | - Features Used:
21 | - `Open`
22 | - `High`
23 | - `Low`
24 | - `Close`
25 | - `Volume`
26 | - The dataset is split into training and testing sets.
27 | - Data is scaled using MinMaxScaler for better model performance.
28 |
29 | ---
30 |
31 | ## 🔧 Technologies Used
32 |
33 | - Python
34 | - NumPy
35 | - Pandas
36 | - Matplotlib / Seaborn
37 | - TensorFlow / Keras
38 | - Scikit-learn
39 |
40 | ---
41 |
42 | ## 🚀 How to Run the Project
43 |
44 | 1. **Clone the Repository**
45 |
46 | ```bash
47 | git clone https://github.com/webvokess/Google-Stock-Price-Prediction-using-LSTM.git
48 | cd Google-Stock-Price-Prediction-using-LSTM
49 |
50 | 2. **Install Required Libraries**
51 | ```bash
52 | pip install -r requirements.txt
53 |
54 | 3. **Run the Notebook**
55 | ```bash
56 | pip install numpy pandas matplotlib seaborn scikit-learn tensorflow
57 |
58 | ## 📈 Model Architecture
59 |
60 | The LSTM model consists of the following layers:
61 |
62 | - **Input Layer:** Preprocessed stock price sequences
63 | - **LSTM Layer 1:** Captures temporal patterns in the data
64 | - **LSTM Layer 2 (optional):** Stacked LSTM to deepen learning
65 | - **Dropout Layers:** Prevent overfitting
66 | - **Dense Layer:** Single output neuron to predict the next closing price
67 |
68 | **Model Configuration:**
69 |
70 | - **Loss Function:** Mean Squared Error (MSE)
71 | - **Optimizer:** Adam
72 | - **Activation Functions:** `tanh` in LSTM, linear in output
73 |
74 | ---
75 |
76 | ## 📌 Results
77 |
78 | - The model was able to **predict Google's stock closing prices** with a trend closely following actual values.
79 | - Visualization typically shows:
80 | - **Blue Line:** Real historical prices
81 | - **Red/Orange Line:** Predicted prices
82 |
83 | ### Key Takeaways:
84 | - LSTM effectively learns temporal dependencies in stock price data.
85 | - Predictions help in understanding potential future price movements.
86 |
87 | ---
88 |
89 | ## 📉 Limitations
90 |
91 | - The model **does not include** external factors such as news, market sentiment, or economic indicators.
92 | - It uses only **historical price data**, which limits real-world forecasting accuracy.
93 | - Real-time or frequent retraining is required for production use.
94 | - No technical indicators (e.g., RSI, MACD) are included — adding these could improve performance.
95 |
96 |
--------------------------------------------------------------------------------
/Stock_price_predection.ipynb:
--------------------------------------------------------------------------------
1 | {
2 | "cells": [
3 | {
4 | "cell_type": "markdown",
5 | "id": "028fc3b4",
6 | "metadata": {},
7 | "source": [
8 | "# Google stock Price prediction using LSTM"
9 | ]
10 | },
11 | {
12 | "cell_type": "code",
13 | "execution_count": 56,
14 | "id": "5f057c9f",
15 | "metadata": {},
16 | "outputs": [],
17 | "source": [
18 | "import numpy as np\n",
19 | "import matplotlib.pyplot as plt\n",
20 | "import pandas as pd\n",
21 | "from sklearn.preprocessing import MinMaxScaler\n",
22 | "from tensorflow.keras.models import Sequential\n",
23 | "from tensorflow.keras.layers import LSTM\n",
24 | "from tensorflow.keras.layers import Dense\n",
25 | "from tensorflow.keras.layers import Dropout"
26 | ]
27 | },
28 | {
29 | "cell_type": "code",
30 | "execution_count": 57,
31 | "id": "81e94c4d",
32 | "metadata": {},
33 | "outputs": [
34 | {
35 | "data": {
36 | "text/html": [
37 | "\n",
38 | "\n",
51 | "
\n",
52 | " \n",
53 | " \n",
54 | " | \n",
55 | " Date | \n",
56 | " Open | \n",
57 | " High | \n",
58 | " Low | \n",
59 | " Close | \n",
60 | " Volume | \n",
61 | "
\n",
62 | " \n",
63 | " \n",
64 | " \n",
65 | " | 0 | \n",
66 | " 1/3/2012 | \n",
67 | " 325.25 | \n",
68 | " 332.83 | \n",
69 | " 324.97 | \n",
70 | " 663.59 | \n",
71 | " 7,380,500 | \n",
72 | "
\n",
73 | " \n",
74 | " | 1 | \n",
75 | " 1/4/2012 | \n",
76 | " 331.27 | \n",
77 | " 333.87 | \n",
78 | " 329.08 | \n",
79 | " 666.45 | \n",
80 | " 5,749,400 | \n",
81 | "
\n",
82 | " \n",
83 | " | 2 | \n",
84 | " 1/5/2012 | \n",
85 | " 329.83 | \n",
86 | " 330.75 | \n",
87 | " 326.89 | \n",
88 | " 657.21 | \n",
89 | " 6,590,300 | \n",
90 | "
\n",
91 | " \n",
92 | " | 3 | \n",
93 | " 1/6/2012 | \n",
94 | " 328.34 | \n",
95 | " 328.77 | \n",
96 | " 323.68 | \n",
97 | " 648.24 | \n",
98 | " 5,405,900 | \n",
99 | "
\n",
100 | " \n",
101 | " | 4 | \n",
102 | " 1/9/2012 | \n",
103 | " 322.04 | \n",
104 | " 322.29 | \n",
105 | " 309.46 | \n",
106 | " 620.76 | \n",
107 | " 11,688,800 | \n",
108 | "
\n",
109 | " \n",
110 | "
\n",
111 | "