├── QuantumPerceptron.pptx
├── qsharp-quick-reference.pdf
├── QuantumPerceptron
├── QuantumPerceptron.csproj
├── QuantumPerceptron.sln
├── ClassicalDriver.cs
├── QuantumClassifier_Easy.qs
└── QuantumClassifier.qs
├── LICENSE
├── README.md
└── .gitignore
/QuantumPerceptron.pptx:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/microsoft/MLADS2018-QuantumML/HEAD/QuantumPerceptron.pptx
--------------------------------------------------------------------------------
/qsharp-quick-reference.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/microsoft/MLADS2018-QuantumML/HEAD/qsharp-quick-reference.pdf
--------------------------------------------------------------------------------
/QuantumPerceptron/QuantumPerceptron.csproj:
--------------------------------------------------------------------------------
1 |
2 |
3 | Exe
4 | netcoreapp2.1
5 | x64
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
--------------------------------------------------------------------------------
/QuantumPerceptron/QuantumPerceptron.sln:
--------------------------------------------------------------------------------
1 |
2 | Microsoft Visual Studio Solution File, Format Version 12.00
3 | # Visual Studio 15
4 | VisualStudioVersion = 15.0.27130.2036
5 | MinimumVisualStudioVersion = 10.0.40219.1
6 | Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "QuantumPerceptron", "QuantumPerceptron.csproj", "{8D1D7634-1578-4A53-AAF1-B0FAE6C4B446}"
7 | EndProject
8 | Global
9 | GlobalSection(SolutionConfigurationPlatforms) = preSolution
10 | Debug|Any CPU = Debug|Any CPU
11 | Release|Any CPU = Release|Any CPU
12 | EndGlobalSection
13 | GlobalSection(ProjectConfigurationPlatforms) = postSolution
14 | {8D1D7634-1578-4A53-AAF1-B0FAE6C4B446}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
15 | {8D1D7634-1578-4A53-AAF1-B0FAE6C4B446}.Debug|Any CPU.Build.0 = Debug|Any CPU
16 | {8D1D7634-1578-4A53-AAF1-B0FAE6C4B446}.Release|Any CPU.ActiveCfg = Release|Any CPU
17 | {8D1D7634-1578-4A53-AAF1-B0FAE6C4B446}.Release|Any CPU.Build.0 = Release|Any CPU
18 | EndGlobalSection
19 | GlobalSection(SolutionProperties) = preSolution
20 | HideSolutionNode = FALSE
21 | EndGlobalSection
22 | GlobalSection(ExtensibilityGlobals) = postSolution
23 | SolutionGuid = {C73A1598-2A3E-4A2E-9EDC-B4CE8B3CDB2F}
24 | EndGlobalSection
25 | EndGlobal
26 |
--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
1 | MIT License
2 |
3 | Copyright (c) Microsoft Corporation. All rights reserved.
4 |
5 | Permission is hereby granted, free of charge, to any person obtaining a copy
6 | of this software and associated documentation files (the "Software"), to deal
7 | in the Software without restriction, including without limitation the rights
8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 |
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 |
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE
22 |
--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | ***This repository has been archived. Please refer to [the tutorial on quantum classification with Q#](https://github.com/microsoft/QuantumKatas/tree/main/tutorials/QuantumClassification) for the latest version of the tutorial.***
2 |
3 | # Welcome!
4 |
5 | This repository contains the materials for the "Introduction to Quantum Machine Learning" workshop.
6 |
7 | In this workshop the participants will get hands-on experience implementing the simplest quantum machine learning algorithm - a quantum perceptron.
8 |
9 | ## Installing and Getting Started
10 |
11 | To work on this tutorial, you'll need to install the [Quantum Development Kit](https://docs.microsoft.com/quantum), available for Windows 10, macOS, and for Linux.
12 | Please see the [install guide for the Quantum Development Kit](https://docs.microsoft.com/quantum/install-guide/) for the detailed instructions. We recommend that you use Visual Studio 2017 or Visual Studio Code.
13 |
14 | If you have Git installed, go on and clone the Microsoft/MLADS2018-QuantumML repository. From your favorite command line:
15 |
16 | ```bash
17 | git clone https://github.com/Microsoft/MLADS2018-QuantumML.git
18 | ```
19 |
20 | > **TIP**: Both Visual Studio 2017 and Visual Studio Code make it easy to clone repositories from within your development environment.
21 | > See the [Visual Studio 2017](https://docs.microsoft.com/en-us/vsts/git/tutorial/clone?view=vsts&tabs=visual-studio#clone-from-another-git-provider) and [Visual Studio Code](https://code.visualstudio.com/docs/editor/versioncontrol#_cloning-a-repository) documentation for details.
22 |
23 | If you don't have Git installed, you can manually download a [standalone copy of the tutorials](https://github.com/Microsoft/MLADS2018-QuantumML/archive/master.zip).
24 |
25 | ## Tutorial Structure
26 |
27 | The tutorial contains the template of the algorithm which you will work on and the classical harness used for running your code. The project is laid out as below.
28 |
29 | ```
30 | README.md # Tutorial instructions
31 | QuantumPerceptron/
32 | QuantumPerceptron.sln # Visual Studio 2017 solution file.
33 | QuantumPerceptron.csproj # Project file used to build both classical and quantum code.
34 |
35 | QuantumClassifier.qs # Q# source code containing the template of the quantum perceptron.
36 | QuantumClassifier_Easy.qs # Q# source code containing the implementations of the quantum perceptron.
37 | ClassicalDriver.cs # C# source code used to load the data, invoke the Q# code and do classical processing.
38 | ```
39 |
40 | To open the tutorial in Visual Studio 2017, open the `QuantumPerceptron.sln` solution file.
41 |
42 | To open the tutorial in Visual Studio Code, open the `QuantumPerceptron/` folder.
43 | Press Ctrl + Shift + P / ⌘ + Shift + P to open the Command Palette and type "Open Folder" on Windows 10 or Linux or "Open" on macOS.
44 |
45 | > **TIP**: Almost all commands available in Visual Studio Code can be found in the Command Palette.
46 | > If you ever get stuck, press Ctrl + Shfit + P / ⌘ + Shift + P and type some letters to search through all available commands.
47 |
48 | > **TIP**: You can also launch Visual Studio Code from the command line if you prefer:
49 | > ```bash
50 | > code QuantumPerceptron/
51 | > ```
52 |
53 | ## Working on the Tutorial
54 |
55 | Once you have the project open, you can build it and run it (`F5` in Visual Studio, `dotnet run` in Visual Studio Code terminal or command line).
56 |
57 | Initially the quantum code does nothing and always reports classification success rate of -1.
58 | Once you fill in the correct code in `QuantumClassifier.qs` file, the code will learn the model parameter which provides the best separation of the given classes.
59 | If you get stuck or run out of time and just want to see the model train,
60 | `QuantumClassifier_Easy.qs` file contains the full code for model training. You can look up the place in which you got stuck or switch to running the code in that file.
61 |
62 | As a stretch goal, try to implement the classical harness and the quantum classification circuit to generate new data and classify it using your trained model.
63 |
64 | ## Useful Links
65 |
66 | * You can find Q# language quick reference [here](./qsharp-quick-reference.pdf).
67 |
68 |
69 | * To improve your experience in the workshop, we suggest you to go through the BasicGates kata from the [Quantum Katas](https://github.com/Microsoft/QuantumKatas) project.
70 | It teaches you the basic gates used in quantum computing and helps you get more comfortable with Q#.
71 | Superposition and Measurement katas are the next best follow-up topics.
72 |
73 | * You can find the slides used in the introductory part of the workshop [here](./QuantumPerceptron.pptx).
74 |
75 | ---
76 |
77 | ## Contributing
78 |
79 | This project welcomes contributions and suggestions. Most contributions require you to agree to a
80 | Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us
81 | the rights to use your contribution. For details, visit https://cla.microsoft.com.
82 |
83 | When you submit a pull request, a CLA-bot will automatically determine whether you need to provide
84 | a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions
85 | provided by the bot. You will only need to do this once across all repos using our CLA.
86 |
87 | This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
88 | For more information see the [Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/) or
89 | contact [opencode@microsoft.com](mailto:opencode@microsoft.com) with any additional questions or comments.
90 |
--------------------------------------------------------------------------------
/.gitignore:
--------------------------------------------------------------------------------
1 | ## Ignore Visual Studio temporary files, build results, and
2 | ## files generated by popular Visual Studio add-ons.
3 | ##
4 | ## Get latest from https://github.com/github/gitignore/blob/master/VisualStudio.gitignore
5 |
6 | # User-specific files
7 | *.suo
8 | *.user
9 | *.userosscache
10 | *.sln.docstates
11 |
12 | # User-specific files (MonoDevelop/Xamarin Studio)
13 | *.userprefs
14 |
15 | # Build results
16 | [Dd]ebug/
17 | [Dd]ebugPublic/
18 | [Rr]elease/
19 | [Rr]eleases/
20 | x64/
21 | x86/
22 | bld/
23 | [Bb]in/
24 | [Oo]bj/
25 | [Ll]og/
26 |
27 | # Visual Studio 2015/2017 cache/options directory
28 | .vs/
29 | # Uncomment if you have tasks that create the project's static files in wwwroot
30 | #wwwroot/
31 |
32 | # Visual Studio 2017 auto generated files
33 | Generated\ Files/
34 |
35 | # MSTest test Results
36 | [Tt]est[Rr]esult*/
37 | [Bb]uild[Ll]og.*
38 |
39 | # NUNIT
40 | *.VisualState.xml
41 | TestResult.xml
42 |
43 | # Build Results of an ATL Project
44 | [Dd]ebugPS/
45 | [Rr]eleasePS/
46 | dlldata.c
47 |
48 | # Benchmark Results
49 | BenchmarkDotNet.Artifacts/
50 |
51 | # .NET Core
52 | project.lock.json
53 | project.fragment.lock.json
54 | artifacts/
55 | **/Properties/launchSettings.json
56 |
57 | # StyleCop
58 | StyleCopReport.xml
59 |
60 | # Files built by Visual Studio
61 | *_i.c
62 | *_p.c
63 | *_i.h
64 | *.ilk
65 | *.meta
66 | *.obj
67 | *.iobj
68 | *.pch
69 | *.pdb
70 | *.ipdb
71 | *.pgc
72 | *.pgd
73 | *.rsp
74 | *.sbr
75 | *.tlb
76 | *.tli
77 | *.tlh
78 | *.tmp
79 | *.tmp_proj
80 | *.log
81 | *.vspscc
82 | *.vssscc
83 | .builds
84 | *.pidb
85 | *.svclog
86 | *.scc
87 |
88 | # Chutzpah Test files
89 | _Chutzpah*
90 |
91 | # Visual C++ cache files
92 | ipch/
93 | *.aps
94 | *.ncb
95 | *.opendb
96 | *.opensdf
97 | *.sdf
98 | *.cachefile
99 | *.VC.db
100 | *.VC.VC.opendb
101 |
102 | # Visual Studio profiler
103 | *.psess
104 | *.vsp
105 | *.vspx
106 | *.sap
107 |
108 | # Visual Studio Trace Files
109 | *.e2e
110 |
111 | # TFS 2012 Local Workspace
112 | $tf/
113 |
114 | # Guidance Automation Toolkit
115 | *.gpState
116 |
117 | # ReSharper is a .NET coding add-in
118 | _ReSharper*/
119 | *.[Rr]e[Ss]harper
120 | *.DotSettings.user
121 |
122 | # JustCode is a .NET coding add-in
123 | .JustCode
124 |
125 | # TeamCity is a build add-in
126 | _TeamCity*
127 |
128 | # DotCover is a Code Coverage Tool
129 | *.dotCover
130 |
131 | # AxoCover is a Code Coverage Tool
132 | .axoCover/*
133 | !.axoCover/settings.json
134 |
135 | # Visual Studio code coverage results
136 | *.coverage
137 | *.coveragexml
138 |
139 | # NCrunch
140 | _NCrunch_*
141 | .*crunch*.local.xml
142 | nCrunchTemp_*
143 |
144 | # MightyMoose
145 | *.mm.*
146 | AutoTest.Net/
147 |
148 | # Web workbench (sass)
149 | .sass-cache/
150 |
151 | # Installshield output folder
152 | [Ee]xpress/
153 |
154 | # DocProject is a documentation generator add-in
155 | DocProject/buildhelp/
156 | DocProject/Help/*.HxT
157 | DocProject/Help/*.HxC
158 | DocProject/Help/*.hhc
159 | DocProject/Help/*.hhk
160 | DocProject/Help/*.hhp
161 | DocProject/Help/Html2
162 | DocProject/Help/html
163 |
164 | # Click-Once directory
165 | publish/
166 |
167 | # Publish Web Output
168 | *.[Pp]ublish.xml
169 | *.azurePubxml
170 | # Note: Comment the next line if you want to checkin your web deploy settings,
171 | # but database connection strings (with potential passwords) will be unencrypted
172 | *.pubxml
173 | *.publishproj
174 |
175 | # Microsoft Azure Web App publish settings. Comment the next line if you want to
176 | # checkin your Azure Web App publish settings, but sensitive information contained
177 | # in these scripts will be unencrypted
178 | PublishScripts/
179 |
180 | # NuGet Packages
181 | *.nupkg
182 | # The packages folder can be ignored because of Package Restore
183 | **/[Pp]ackages/*
184 | # except build/, which is used as an MSBuild target.
185 | !**/[Pp]ackages/build/
186 | # Uncomment if necessary however generally it will be regenerated when needed
187 | #!**/[Pp]ackages/repositories.config
188 | # NuGet v3's project.json files produces more ignorable files
189 | *.nuget.props
190 | *.nuget.targets
191 |
192 | # Microsoft Azure Build Output
193 | csx/
194 | *.build.csdef
195 |
196 | # Microsoft Azure Emulator
197 | ecf/
198 | rcf/
199 |
200 | # Windows Store app package directories and files
201 | AppPackages/
202 | BundleArtifacts/
203 | Package.StoreAssociation.xml
204 | _pkginfo.txt
205 | *.appx
206 |
207 | # Visual Studio cache files
208 | # files ending in .cache can be ignored
209 | *.[Cc]ache
210 | # but keep track of directories ending in .cache
211 | !*.[Cc]ache/
212 |
213 | # Others
214 | ClientBin/
215 | ~$*
216 | *~
217 | *.dbmdl
218 | *.dbproj.schemaview
219 | *.jfm
220 | *.pfx
221 | *.publishsettings
222 | orleans.codegen.cs
223 |
224 | # Including strong name files can present a security risk
225 | # (https://github.com/github/gitignore/pull/2483#issue-259490424)
226 | #*.snk
227 |
228 | # Since there are multiple workflows, uncomment next line to ignore bower_components
229 | # (https://github.com/github/gitignore/pull/1529#issuecomment-104372622)
230 | #bower_components/
231 |
232 | # RIA/Silverlight projects
233 | Generated_Code/
234 |
235 | # Backup & report files from converting an old project file
236 | # to a newer Visual Studio version. Backup files are not needed,
237 | # because we have git ;-)
238 | _UpgradeReport_Files/
239 | Backup*/
240 | UpgradeLog*.XML
241 | UpgradeLog*.htm
242 | ServiceFabricBackup/
243 | *.rptproj.bak
244 |
245 | # SQL Server files
246 | *.mdf
247 | *.ldf
248 | *.ndf
249 |
250 | # Business Intelligence projects
251 | *.rdl.data
252 | *.bim.layout
253 | *.bim_*.settings
254 | *.rptproj.rsuser
255 |
256 | # Microsoft Fakes
257 | FakesAssemblies/
258 |
259 | # GhostDoc plugin setting file
260 | *.GhostDoc.xml
261 |
262 | # Node.js Tools for Visual Studio
263 | .ntvs_analysis.dat
264 | node_modules/
265 |
266 | # Visual Studio 6 build log
267 | *.plg
268 |
269 | # Visual Studio 6 workspace options file
270 | *.opt
271 |
272 | # Visual Studio 6 auto-generated workspace file (contains which files were open etc.)
273 | *.vbw
274 |
275 | # Visual Studio LightSwitch build output
276 | **/*.HTMLClient/GeneratedArtifacts
277 | **/*.DesktopClient/GeneratedArtifacts
278 | **/*.DesktopClient/ModelManifest.xml
279 | **/*.Server/GeneratedArtifacts
280 | **/*.Server/ModelManifest.xml
281 | _Pvt_Extensions
282 |
283 | # Paket dependency manager
284 | .paket/paket.exe
285 | paket-files/
286 |
287 | # FAKE - F# Make
288 | .fake/
289 |
290 | # JetBrains Rider
291 | .idea/
292 | *.sln.iml
293 |
294 | # CodeRush
295 | .cr/
296 |
297 | # Python Tools for Visual Studio (PTVS)
298 | __pycache__/
299 | *.pyc
300 |
301 | # Cake - Uncomment if you are using it
302 | # tools/**
303 | # !tools/packages.config
304 |
305 | # Tabs Studio
306 | *.tss
307 |
308 | # Telerik's JustMock configuration file
309 | *.jmconfig
310 |
311 | # BizTalk build output
312 | *.btp.cs
313 | *.btm.cs
314 | *.odx.cs
315 | *.xsd.cs
316 |
317 | # OpenCover UI analysis results
318 | OpenCover/
319 |
320 | # Azure Stream Analytics local run output
321 | ASALocalRun/
322 |
323 | # MSBuild Binary and Structured Log
324 | *.binlog
325 |
326 | # NVidia Nsight GPU debugger configuration file
327 | *.nvuser
328 |
329 | # MFractors (Xamarin productivity tool) working folder
330 | .mfractor/
331 |
--------------------------------------------------------------------------------
/QuantumPerceptron/ClassicalDriver.cs:
--------------------------------------------------------------------------------
1 | // Copyright (c) Microsoft Corporation. All rights reserved.
2 | // Licensed under the MIT license.
3 |
4 | ////////////////////////////////////////////////////////////////////////////
5 | // This file contains the classical code of the quantum perceptron
6 | // which loads the data, calls quantum classification routines
7 | // and performs search for model parameters.
8 | ////////////////////////////////////////////////////////////////////////////
9 |
10 | using Microsoft.Quantum.Simulation.Core;
11 | using Microsoft.Quantum.Simulation.Simulators;
12 | using System;
13 |
14 | namespace Microsoft.Quantum.MachineLearning
15 | {
16 | class ClassicalDriver
17 | {
18 | // Call quantum classification on the given training data for the given model parameter
19 | // and return its success rate.
20 | static double GetClassificationSuccessRate(double[] data, long[] labels, double angle)
21 | {
22 | double s = 0;
23 | using (var qsim = new QuantumSimulator(true, 123))
24 | {
25 | // Change the following line to call QuantumClassifier_SuccessRate_Easy.Run to run pre-written model training code
26 | s = QuantumClassifier_SuccessRate.Run(qsim, angle, new QArray(data), new QArray(labels)).Result;
27 | }
28 | return s;
29 | }
30 |
31 | static void SaveAsCsv(double[] data, long[] labels)
32 | {
33 | string fileContents = "Label,Data" + Environment.NewLine;
34 | for (int i = 0; i < data.Length; ++i)
35 | {
36 | fileContents += labels[i] + "," + data[i].ToString("F4") + Environment.NewLine;
37 | }
38 | System.IO.File.WriteAllText("trainingData.csv", fileContents);
39 | }
40 |
41 | ///
42 | /// Create a data set in the range [0, 2π] with a gap of the given size at the given central angle
43 | /// and a symmetrical gap at the angle π + the given central angle.
44 | ///
45 | /// Number of random data points
46 | /// Where to create the gap
47 | /// Width of gap
48 | ///
49 | static double[] CreateSeparableDataset(int size, double center, double gap)
50 | {
51 | double[] data = new double[size];
52 | Random rnd = new Random(123);
53 |
54 | double max = Math.PI;
55 |
56 | // These variables are used to create the gap
57 | double lowerMax = center - gap / 2.0;
58 | double upperScale = (max - center - gap / 2.0) / (max - center);
59 |
60 | for (int i = 0; i < size; i++)
61 | {
62 | // 1. Generate a random number without taking the gap into account
63 | data[i] = rnd.NextDouble() * max;
64 |
65 | // 2. Shift it to create the gap
66 | if (data[i] <= center)
67 | {
68 | // This is the simple case, just re-scale proportionally (3-simple rule)
69 | data[i] *= lowerMax / center;
70 | }
71 | else
72 | {
73 | // Slightly more complicated: invert, 3-simple rule to rescale, invert again
74 | data[i] = max - data[i];
75 | data[i] *= upperScale;
76 | data[i] = max - data[i];
77 | }
78 |
79 | // 3. Spread to the full circle (2π) -- we'll have two gaps
80 | if (rnd.Next(2) == 1)
81 | {
82 | data[i] += Math.PI;
83 | }
84 | }
85 |
86 | return data;
87 | }
88 |
89 | static void Main(string[] args)
90 | {
91 | // Larger dataset size gives more precise training result but slower training process
92 | int datasetSize = 200;
93 |
94 | // Prepare a perfectly separable training dataset
95 | double separationAngle = 2.0;
96 | double correctAngle = separationAngle - Math.PI / 2;
97 | double margin = 0.1;
98 |
99 | double[] data = CreateSeparableDataset(datasetSize, separationAngle, margin);
100 |
101 | // Assign labels to training data
102 | long[] labels = new long[datasetSize];
103 | for (int i = 0; i < datasetSize; ++i)
104 | {
105 | labels[i] = (data[i] > separationAngle && data[i] < separationAngle + Math.PI ? 1 : 0);
106 | }
107 |
108 | // Save the training data to file
109 | // SaveAsCsv(data, labels);
110 |
111 | // Check classification success rate on the angle that we know to be correct
112 | Console.Out.WriteLine($"For correct angle {correctAngle.ToString("F3")} success rate = " +
113 | GetClassificationSuccessRate(data, labels, correctAngle));
114 |
115 | long msStart = DateTime.Now.Ticks / TimeSpan.TicksPerMillisecond;
116 | // Do a ternary search on the classification angle (we know that the angle is between 0 and 2*PI).
117 | // Normally we would apply a better technique, like gradient search, but a demo which runs fast enough lacks precision for those techniques.
118 | double l = 0.0, r = 2 * Math.PI;
119 | int iter = 0;
120 | while (Math.Abs(l-r) > margin) // stop when the length of the search interval is smaller than the margin
121 | {
122 | iter++;
123 | double m1 = l + (r - l) / 3;
124 | double m2 = r - (r - l) / 3;
125 | double s1 = GetClassificationSuccessRate(data, labels, m1);
126 | double s2 = GetClassificationSuccessRate(data, labels, m2);
127 | if (s1 <= s2)
128 | {
129 | l = m1;
130 | } else
131 | {
132 | r = m2;
133 | }
134 | Console.Out.WriteLine($"Iteration {iter}: angle {m1.ToString("F3")} -> {s1}, angle {m2.ToString("F3")} -> {s2}, narrowing search to [{l.ToString("F3")}, {r.ToString("F3")}]");
135 | }
136 |
137 | // Report training results (learned model parameter and classification success rate for it)
138 | double s = GetClassificationSuccessRate(data, labels, (l+r)/2);
139 | Console.Out.WriteLine($"Training result: angle {((l+r)/2).ToString("F3")} -> {s}");
140 | long msEnd = DateTime.Now.Ticks / TimeSpan.TicksPerMillisecond;
141 | Console.Out.WriteLine("Training time " + (msEnd - msStart) / 1000.0);
142 |
143 | Console.WriteLine("Press any key to continue...");
144 | Console.ReadKey();
145 |
146 | ////////////////////////////////////////////////////////////////////////
147 | // Stretch goal: Write your own classifier!
148 | //
149 | // Now that you've trained your model, it's time to see how it performs on new data.
150 | // Generate new data following the same process as for the training data.
151 | // Write classification circuit QuantumClassifier following the suggestions in QuantumClassifier.qs.
152 | // Run it and compare classification results to the labels you assign when generating the data.
153 | ////////////////////////////////////////////////////////////////////////
154 | }
155 | }
156 | }
157 |
--------------------------------------------------------------------------------
/QuantumPerceptron/QuantumClassifier_Easy.qs:
--------------------------------------------------------------------------------
1 | // Copyright (c) Microsoft Corporation. All rights reserved.
2 | // Licensed under the MIT license.
3 |
4 | ////////////////////////////////////////////////////////////////////////////
5 | // This file contains the easy version of the workshop.
6 | // All the quantum code is already written, and you can run the program to see how it works.
7 | // To switch to this version, replace QuantumClassifier_SuccessRate in ClassicalDriver.cs
8 | // with QuantumClassifier_SuccessRate_Easy.
9 | //
10 | // We recommend you to try the harder version (QuantumClassifier.qs file),
11 | // where you have to write Q# model training code following prompts in the comments,
12 | // and eventually write the classifier code on your own.
13 | ////////////////////////////////////////////////////////////////////////////
14 |
15 | namespace Microsoft.Quantum.MachineLearning {
16 |
17 | open Microsoft.Quantum.Primitive;
18 | open Microsoft.Quantum.Canon;
19 | open Microsoft.Quantum.Extensions.Convert;
20 |
21 |
22 | ////////////////////////////////////////////////////////////////////////
23 | // Encoding circuit
24 | //
25 | // # Summary
26 | // Prepares two qubits in a state which represents a point in the training dataset.
27 | // # Inputs
28 | // data : The input vector (single floating-point number)
29 | // label : The input label (0 or 1)
30 | // dataQubit : The qubit to be prepared in a state which represents input vector
31 | // labelQubit : The qubit to be prepared in a state which represents the label
32 | // (|0⟩ or |1⟩ for labels 0 or 1, respectively)
33 | ////////////////////////////////////////////////////////////////////////
34 | operation EncodeDataInQubits_Easy (
35 | data : Double,
36 | label : Int,
37 | dataQubit : Qubit,
38 | labelQubit : Qubit) : Unit {
39 |
40 | // Make sure both qubits start in |0⟩ state (you can use library operation Reset)
41 | Reset(dataQubit);
42 | Reset(labelQubit);
43 |
44 | // Encode the input vector in dataQubit state using Ry rotation gate.
45 | // Note that the rotation angle has to be exactly "data" to be consistent with the labels generation in Driver.cs
46 | Ry(data, dataQubit);
47 |
48 | // Encode the label in labelQubit state: |0⟩ or |1⟩ for labels 0 or 1 (you can use X gate to change qubit state |0⟩ to |1⟩)
49 | if (label == 1) {
50 | X(labelQubit);
51 | }
52 | }
53 |
54 |
55 | ////////////////////////////////////////////////////////////////////////
56 | // Single-shot single-point validation circuit
57 | //
58 | // # Summary
59 | // Classifies a data point encoded as a qubit and validates the result against the expected label.
60 | // # Inputs
61 | // alpha : The model parameter used for classification
62 | // dataQubit : The qubit which represents the input state
63 | // labelQubit : The qubit which represents the expected label of the input state
64 | // # Result
65 | // True if the data point has been classified correctly,
66 | // false if it has been misclassified.
67 | //
68 | // In classification scenario, the circuit will be the same, but the label qubit will always start in |0⟩ state,
69 | // and to get the classification result you'll measure it.
70 | ////////////////////////////////////////////////////////////////////////
71 | operation Validate_Easy (
72 | alpha : Double,
73 | dataQubit : Qubit,
74 | labelQubit : Qubit) : Bool {
75 |
76 | // Rotate the state of the data qubit by -alpha;
77 | // this will get it close to the |0⟩ state if the data point belonged to class 0,
78 | // and to the |1⟩ state if the data point belonged to class 1
79 | Ry(-alpha, dataQubit);
80 |
81 | // Apply CNOT with data qubit as control and label qubit as target
82 | // to compute XOR of the expected label and the computed label on labelQubit
83 | CNOT(dataQubit, labelQubit);
84 |
85 | // Measure the label qubit in computational basis
86 |
87 | // If the measurement result is |0⟩, XOR of the expected label and the computed label is 0,
88 | // which means that the labels are the same and classification was correct.
89 | // Return the classification success
90 | return M(labelQubit) == Zero;
91 | }
92 |
93 |
94 | ////////////////////////////////////////////////////////////////////////
95 | // Full validation routine
96 | //
97 | // # Summary
98 | // Given the value of the model parameter (average rotation angle of the data points in class 0),
99 | // # Inputs
100 | // alpha : The model parameter used for classification
101 | // dataPoints : An array of training vectors (individual floating-point numbers)
102 | // labels : An array of training labels (0 or 1)
103 | // # Result
104 | // The success rate of classification for the given model parameter.
105 | ////////////////////////////////////////////////////////////////////////
106 | operation QuantumClassifier_SuccessRate_Easy (
107 | alpha : Double,
108 | dataPoints : Double[],
109 | labels : Int[]) : Double {
110 |
111 | let N = Length(dataPoints);
112 | // The number of times each point is classified; larger values give higher accuracy but longer run time
113 | let nSamples = 201;
114 | // Define a mutable variable to store the number of correctly classified points in the dataset
115 | mutable nCorrectPoints = 0;
116 |
117 | // Allocate two qubits to be used in the classification
118 | using ((dataQubit, labelQubit) = (Qubit(), Qubit())) {
119 |
120 | // Iterate over all points of the dataset
121 | for (i in 0 .. N - 1) {
122 |
123 | // Define a mutable variable to store the number of successful classification runs
124 | mutable nCorrectClassificationRuns = 0;
125 |
126 | // Classify i-th data point by running classification circuit nSamples times
127 | for (j in 1 .. nSamples) {
128 | // Prepare data qubit and label qubit in a state which encodes the j-th data point
129 | EncodeDataInQubits_Easy(dataPoints[i], labels[i], dataQubit, labelQubit);
130 |
131 | // Run classification on the prepared qubits and count the runs when it succeeded
132 | if (Validate_Easy(alpha, dataQubit, labelQubit)) {
133 | set nCorrectClassificationRuns = nCorrectClassificationRuns + 1;
134 | }
135 | }
136 |
137 | // The point in the dataset has been classified correctly if
138 | // the share of runs on which classification succeeded is greater than 50%.
139 | if (nCorrectClassificationRuns * 2 > nSamples) {
140 | set nCorrectPoints = nCorrectPoints + 1;
141 | }
142 | }
143 |
144 | // Clean up both qubits before deallocating them using library operation Reset.
145 | Reset(dataQubit);
146 | Reset(labelQubit);
147 | }
148 |
149 | // Return the success rate of the classification (the percentage of points that have been classified correctly)
150 | // Note that you need ToDouble library function to convert integer numbers to doubles explicitly
151 | return ToDouble(nCorrectPoints) / ToDouble(N);
152 | }
153 | }
154 |
--------------------------------------------------------------------------------
/QuantumPerceptron/QuantumClassifier.qs:
--------------------------------------------------------------------------------
1 | // Copyright (c) Microsoft Corporation. All rights reserved.
2 | // Licensed under the MIT license.
3 |
4 | ////////////////////////////////////////////////////////////////////////////
5 | // This file contains the main version of the workshop.
6 | // You have to write Q# model training code following prompts in the comments,
7 | // and eventually write the classifier code on your own.
8 | //
9 | // If you get stuck on some syntax or run out of time and just want to see the model train,
10 | // take a look at the QuantumClassifier_Easy.qs file - it contains the full code
11 | // for model training.
12 | ////////////////////////////////////////////////////////////////////////////
13 |
14 | namespace Microsoft.Quantum.MachineLearning {
15 |
16 | open Microsoft.Quantum.Primitive;
17 | open Microsoft.Quantum.Canon;
18 | open Microsoft.Quantum.Extensions.Convert;
19 |
20 |
21 | ////////////////////////////////////////////////////////////////////////
22 | // Encoding circuit
23 | //
24 | // # Summary
25 | // Prepares two qubits in a state which represents a point in the training dataset.
26 | // # Inputs
27 | // data : The input vector (single floating-point number)
28 | // label : The input label (0 or 1)
29 | // dataQubit : The qubit to be prepared in a state which represents input vector
30 | // labelQubit : The qubit to be prepared in a state which represents the label
31 | // (|0⟩ or |1⟩ for labels 0 or 1, respectively)
32 | ////////////////////////////////////////////////////////////////////////
33 | operation EncodeDataInQubits (
34 | data : Double,
35 | label : Int,
36 | dataQubit : Qubit,
37 | labelQubit : Qubit) : Unit {
38 |
39 | // Make sure both qubits start in |0⟩ state (you can use library operation Reset)
40 | // ...
41 |
42 | // Encode the input vector in dataQubit state using Ry rotation gate.
43 | // Note that the rotation angle has to be exactly "data" to be consistent with the labels generation in Driver.cs
44 | // ...
45 |
46 | // Encode the label in labelQubit state: |0⟩ or |1⟩ for labels 0 or 1 (you can use X gate to change qubit state |0⟩ to |1⟩)
47 | // ...
48 | }
49 |
50 |
51 | ////////////////////////////////////////////////////////////////////////
52 | // Single-shot single-point validation circuit
53 | //
54 | // # Summary
55 | // Classifies a data point encoded as a qubit and validates the result against the expected label.
56 | // # Inputs
57 | // alpha : The model parameter used for classification
58 | // dataQubit : The qubit which represents the input state
59 | // labelQubit : The qubit which represents the expected label of the input state
60 | // # Result
61 | // True if the data point has been classified correctly,
62 | // false if it has been misclassified.
63 | //
64 | // In classification scenario, the circuit will be the same, but the label qubit will always start in |0⟩ state,
65 | // and to get the classification result you'll measure it.
66 | ////////////////////////////////////////////////////////////////////////
67 | operation Validate (
68 | alpha : Double,
69 | dataQubit : Qubit,
70 | labelQubit : Qubit) : Bool {
71 |
72 | // Rotate the state of the data qubit by -alpha;
73 | // this will get it close to the |0⟩ state if the data point belonged to class 0,
74 | // and to the |1⟩ state if the data point belonged to class 1
75 | // ...
76 |
77 | // Apply CNOT with data qubit as control and label qubit as target
78 | // to compute XOR of the expected label and the computed label on labelQubit
79 | // ...
80 |
81 | // Measure the label qubit in computational basis
82 | // ...
83 |
84 | // If the measurement result is |0⟩, XOR of the expected label and the computed label is 0,
85 | // which means that the labels are the same and classification was correct.
86 | // Return the classification success
87 | return false;
88 | }
89 |
90 |
91 | ////////////////////////////////////////////////////////////////////////
92 | // Full validation routine
93 | //
94 | // # Summary
95 | // Given the value of the model parameter (average rotation angle of the data points in class 0),
96 | // # Inputs
97 | // alpha : The model parameter used for classification
98 | // dataPoints : An array of training vectors (individual floating-point numbers)
99 | // labels : An array of training labels (0 or 1)
100 | // # Result
101 | // The success rate of classification for the given model parameter.
102 | ////////////////////////////////////////////////////////////////////////
103 | operation QuantumClassifier_SuccessRate (
104 | alpha : Double,
105 | dataPoints : Double[],
106 | labels : Int[]) : Double {
107 |
108 | let N = Length(dataPoints);
109 | // The number of times each point is classified; larger values give higher accuracy but longer run time
110 | let nSamples = 201;
111 | // Define a mutable variable to store the number of correctly classified points in the dataset
112 | // ...
113 |
114 | // Allocate two qubits to be used in the classification
115 | using ((dataQubit, labelQubit) = (Qubit(), Qubit())) {
116 |
117 | // Iterate over all points of the dataset
118 | for (i in 0 .. N - 1) {
119 |
120 | // Define a mutable variable to store the number of successful classification runs
121 | // ...
122 |
123 | // Classify i-th data point by running classification circuit nSamples times
124 | for (j in 1 .. nSamples) {
125 | // Prepare data qubit and label qubit in a state which encodes the j-th data point
126 | // ...
127 |
128 | // Run classification on the prepared qubits and count the runs when it succeeded
129 | // ...
130 | }
131 |
132 | // The point in the dataset has been classified correctly if
133 | // the share of runs on which classification succeeded is greater than 50%.
134 | // ...
135 | }
136 |
137 | // Clean up both qubits before deallocating them using library operation Reset.
138 | // ...
139 | }
140 |
141 | // Return the success rate of the classification (the percentage of points that have been classified correctly)
142 | // Note that you need ToDouble library function to convert integer numbers to doubles explicitly
143 | return -1.0;
144 | }
145 |
146 | ////////////////////////////////////////////////////////////////////////
147 | // Stretch goal: Write your own classifier!
148 | //
149 | // The classification logic is similar to the validation one:
150 | // to classify a point in a dataset, you encode it in a data qubit,
151 | // run classification circuit on it with label qubit in |0⟩ state and measure the label qubit;
152 | // repeat this enough times and return the classification outcome which you get more frequently.
153 | ////////////////////////////////////////////////////////////////////////
154 | operation QuantumClassifier (
155 | alpha : Double,
156 | dataPoints : Double[]) : Int[] {
157 |
158 | // ...
159 | return new Int[0];
160 | }
161 | }
162 |
163 |
164 |
--------------------------------------------------------------------------------