{"id":29005444,"url":"https://github.com/absarraashid3/fraudscope-","last_synced_at":"2025-06-25T11:33:16.190Z","repository":{"id":301035215,"uuid":"1007962168","full_name":"AbsarRaashid3/FraudScope-","owner":"AbsarRaashid3","description":"FraudScope is a complete machine learning pipeline for detecting fraudulent financial transactions using only classical ML techniques and core concepts. Designed as a hands-on, transparent system for fraud analytics, it guides users through every stage—from raw data to model evaluation—using only scikit-learn and core Python libraries.","archived":false,"fork":false,"pushed_at":"2025-06-24T20:06:30.000Z","size":1270,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-06-24T21:23:44.229Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/AbsarRaashid3.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-06-24T20:03:38.000Z","updated_at":"2025-06-24T20:06:33.000Z","dependencies_parsed_at":"2025-06-24T21:23:52.408Z","dependency_job_id":"dc0869c9-c1da-4224-83e0-263b3be28854","html_url":"https://github.com/AbsarRaashid3/FraudScope-","commit_stats":null,"previous_names":["absarraashid3/fraudscope-"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/AbsarRaashid3/FraudScope-","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AbsarRaashid3%2FFraudScope-","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AbsarRaashid3%2FFraudScope-/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AbsarRaashid3%2FFraudScope-/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AbsarRaashid3%2FFraudScope-/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/AbsarRaashid3","download_url":"https://codeload.github.com/AbsarRaashid3/FraudScope-/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AbsarRaashid3%2FFraudScope-/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":261863502,"owners_count":23221564,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2025-06-25T11:31:17.551Z","updated_at":"2025-06-25T11:33:16.172Z","avatar_url":"https://github.com/AbsarRaashid3.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# FraudScope: Transaction Fraud Detection \u0026 Analysis\nFraudScope is a complete machine learning pipeline for detecting fraudulent financial transactions using only classical ML techniques and core concepts. Designed as a hands-on, transparent system for fraud analytics, it guides users through every stage—from raw data to model evaluation—using only scikit-learn and core Python libraries.\n\n# Dataset \u0026 Problem Statement\nDataset Source: MoMTSim_20240722202413_1000_dataset.csv (synthetic transactional data)\n\nObjective: Build a classifier to detect fraudulent transactions (isFraud)\n\nFeatures: Includes transaction metadata such as:\n\namount, oldBalInitiator, newBalInitiator\n\ntransactionType, origin, destination, etc.\n\n# Architecture Overview\nA fully modular pipeline:\n\n\nRaw CSV Data\n   ↓\nExploratory Data Analysis (EDA)\n   ↓\nHandling Missing Values \u0026 Duplicates\n   ↓\nOutlier Detection (IQR \u0026 Z-score methods)\n   ↓\nEncoding \u0026 Feature Engineering\n   ↓\nFeature Scaling (MinMaxScaler, StandardScaler)\n   ↓\nFeature Selection (Correlation Matrix Filtering)\n   ↓\nTrain/Test Split (80/20, stratified on 'isFraud')\n   ↓\nModel Training \u0026 Hyperparameter Tuning:\n   - K-Nearest Neighbors (varied k, distance metrics)\n   - Support Vector Machine (kernel, C tuning)\n   - Decision Tree (depth, min_samples_split)\n   - Logistic Regression (L1 \u0026 L2 regularization)\n   ↓\nEvaluation:\n   - Accuracy, Precision, Recall, F1-score\n   - Confusion Matrix \u0026 Classification Report\n   - Training \u0026 Inference Time Comparison\n     \n# Setup \u0026 Requirements\nInstall dependencies with:\n\nbash\nCopy\nEdit\npip install -r requirements.txt\nRequired libraries:\n\npandas, numpy\n\nmatplotlib, seaborn\n\nscikit-learn\n\n # Notebook Walkthrough\nfraudscope.ipynb:\n\nEDA: Value counts, missing data, class imbalance\n\nVisualizations: Boxplots, heatmaps, pairplots\n\nOutlier Handling: IQR and Z-score thresholds\n\nFeature Engineering: Encoding transaction types, calculating balance deltas\n\nScaling: Compared MinMax and Z-score scaling effects\n\nModel Training: Manual loop for each model with grid-tuned hyperparameters\n\nEvaluation: Comparative analysis of model metrics and runtime\n\n# Results \u0026 Model Comparisons\nModel\tF1-Score\tAccuracy\tInference Time\nLogistic Regression\t0.92\t0.94\t⚡ Fast\nKNN (k=5, Euclidean)\t0.88\t0.91\t🐢 Slow\nDecision Tree\t0.93\t0.95\t⚡ Fast\nSVM (RBF)\t0.94\t0.96\t⚖️ Moderate\n\nBest performer: SVM (RBF kernel) in terms of F1 and overall accuracy\n\nMost efficient: Logistic Regression (fastest inference)\n\n# Usage \u0026 Reproducibility\nTo run the notebook:\n\nPlace the dataset (MoMTSim_20240722202413_1000_dataset.csv) in the root directory.\n\nOpen and run fraudscope_analysis.ipynb in Jupyter or Google Colab.\n\nFollow the cell-by-cell analysis and observe output metrics.\n\n# Contributing\nFeel free to fork this repo and submit a pull request if you'd like to:\n\nAdd new models (e.g., ensemble methods)\n\nImprove feature engineering\n\nVisualize confusion matrices interactively\n\n# License \u0026 Academic Integrity\nThis project is for educational and demonstrative purposes only. Do not reuse directly in academic submissions unless explicitly permitted.\n\n# Developed by\nAbsar Raashid\nGitHub Repository\nOpen for feedback and collaboration.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fabsarraashid3%2Ffraudscope-","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fabsarraashid3%2Ffraudscope-","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fabsarraashid3%2Ffraudscope-/lists"}