Mini Project II (Research-Oriented Project with Implementation) Notes | B.Sc. (Computer Science) Semester 6 | Mumbai University | munotes
Official Notes munotes.in
Mini Project II (Research-Oriented Project with Implementation)
B.SC. (COMPUTER SCIENCE) · SEMESTER 6
Strictly as per the University of Mumbai NEP syllabus in force for B.Sc. (Computer Science)
For B.Sc. (Computer Science) students of the University of Mumbai and all its affiliated colleges
Open the book ↓munotes.in Third Year
Contents
Module I Research Planning, Literature Survey & Methodology Design Phase
- How Mini Project II Is Marked: the Guide's Twenty, the Examiner's Thirty and Passing Each 1
- Research in Sixty Hours: the Process from a Question to a Paper 8
- The Worked Study in This Book: a Digit Recognizer Small Enough for a Small Device 14
- Research Problem Identification: From an Interest to a Researchable Problem 20
- Selecting a Topic: Sources of Ideas and the Five Tests a Topic Must Pass 25
- Selecting a Topic in Six Domains: AI, Data Science, Cyber Security, Cloud, IoT and Software Engineering 30
- Formulating the Problem Statement 35
- Research Objectives and the Research Questions They Answer 40
- Before the Hypothesis: Variables, Measures and Controls 45
- The Hypothesis: When a Study Needs One and How to Write It 50
- The Literature Review: What It Is For and How It Is Done 55
- Finding Research Papers: Search Services, Keywords and Search Strings 60
- Judging a Source: Peer Review, Preprints, Venues and Predatory Journals 65
- Reading a Research Paper in Three Passes 70
- Recording What You Read: Notes, Zotero and BibTeX 75
- Five to Eight Papers in Practice: the Eight the Worked Study Read 80
- Identifying the Research Gap 88
- Comparative Analysis of the Work Reviewed 94
- Summarising the Existing Techniques 99
- Research Methodology Design: Turning Objectives into an Experiment 105
- Selecting a Dataset: Sources, Licences and Documentation 111
- The Worked Study's Dataset: the UCI Handwritten Digits, Looked at Closely 117
- Collecting Your Own Data: Logs, Sensors, Surveys and Consent under the DPDP Act 123
- Selecting the Algorithm or Model: Matching the Method to the Question 129
- The Worked Study's Six Candidates and What Each One Stores 133
- Selecting the Tools: Language, Libraries, Environment and Versions 139
- Planning the Experimental Setup: Splits, Folds, Seeds and Leakage 143
- Defining the Evaluation Metrics: Accuracy, Precision, Recall, F1 and the Confusion Matrix 148
- Defining the Evaluation Metrics Beyond Accuracy: Size, Time and How Sure We Are 156
- Research Design Documentation: the Four Documents of Module 1 162
- The Research Proposal 167
- The Literature Review Matrix 175
- The System Model and the Architecture Diagram 181
- The Experimental Design Plan 186
- The Module 1 Review: Research Problem Identification and Literature Review Before the Guide 192
Module II Implementation, Experimentation & Evaluation Phase
- Module 2 at a Glance: Implementation, Experimentation and Evaluation 197
- System Implementation: Organising the Project's Code and Results 202
- Model Building: Pipelines, Preprocessing and Fitting 207
- Model Building for the Six Candidates, and Choosing k by Cross-Validation 212
- Algorithm Implementation: Writing k-Nearest Neighbours Yourself 220
- Algorithm Implementation: PCA by Hand, Checked Against scikit-learn 227
- The Experimental Setup in Code: One Protocol, One Results Table 234
- Code Optimization: Measure First, Then Vectorise 240
- Code Optimization: Making the Model Smaller 246
- Methodology and Implementation Progress: What the Guide Checks in Module 2 252
- Experimental Evaluation: Running the Planned Experiment 258
- Model Training and Testing: Ten Folds, Then the Test Set Once 263
- Performance Metric Calculation: From Predictions to Scores, by Hand and by Program 269
- Performance Metric Calculation for Size and Time 278
- Baseline Comparison: Trivial, Simple and Strong Baselines 286
- Baseline Comparison: Is the Difference Real? McNemar's Test and the Paired Test 291
- Result Analysis: Reading the Results Table 298
- Graphical Representation: Choosing and Drawing the Right Chart 304
- Graphical Representation: the Accuracy Against Size Chart 310
- Interpreting the Results: What the Errors Show 316
- Validating the Hypotheses 322
- Identifying the Limitations 329
- Research Paper Writing: the IEEE Conference Paper and Its Parts 336
- Citing and Referencing in IEEE Style 341
- Honest Research: Plagiarism, AI Tools and Reporting What Happened 347
- Writing the Title and the Abstract 353
- Writing the Introduction 358
- Writing the Related Work Section 364
- Writing the Methodology Section 374
- Writing the Results Section: Tables, Figures and Captions 384
- Writing the Conclusion and the Future Scope 397
- The Final Deliverables: What Is Due at the End of Module 2 403
- The Working Prototype 408
- The Research Paper in IEEE Format: Six to Eight Pages 415
- The GitHub Repository 423
- The Presentation 431
- The Internal Seminar / Review 439
- The External Evaluation: How the Thirty Marks Are Earned 445
- The Viva Voce: Questions You Must Be Able to Answer 451
- Publishing or Presenting Your Work 459
- The Whole Project on One Page: a Checklist from Topic to Viva 466
The chapters
Every chapter of this book comes with the B.Sc. (Computer Science) Semester 6 notes.
The cover and the contents are free to look through. Buy the notes once to read every chapter of every subject in this semester.
Notes + practical solutions + solved papers: ₹798 Already bought it? Sign in
Or notes + practical solutions: ₹499
Or solved papers only: ₹499
Free either way: question papers, the syllabus, and the cover and contents of every book.
The rest of this subject
These notes are cut from the University's printed syllabus. Open the syllabus itself for the same subject.