Submission Folder added for DSC

[ViggiXDevp]

Related Files Regarding DSC attached in that folder

Folder Contains

Crime Data EDA & Visualization Project 🧑‍💻📊📈

The Crime Data Analysis project aims on explore and analyse crime records through meaningful patterns. By performing thorough exploratory data analysis (EDA), data cleaning, data processing and feature engineering, the project seeks to identify trends and patterns related to time, locations and crime patterns. Thorough Visualisation through bar graphs, histograms and heatmaps, it provides insights which are helpful for law enforcement agencies in decision making and resource allocation.

---

📋 Selected Task

Task 1: Exploratory Data Analysis (EDA) on crime data including cleaning, preprocessing, feature engineering and visualization.
Task 2: Building a Scikit-learn Pipline including a suitable model with train, test and evaluate using relevant metrics

---

✅ Features Implemented

Task 1

• Data cleaning and preprocessing with pandas
• Feature engineering to discover hidden patterns
• Multiple visualizations: bar charts, line charts, heatmaps
• Interactive HTML report generated using ydata_profiling
• Modular Python scripts for analysis and plotting
• Ready-to-run Jupyter Notebook for step-by-step execution

Task 2

• Built a Scikit-learn pipeline including preprocessing, model training, testing, and evaluation with relevant metrics
• Applied appropriate machine learning models for crime pattern prediction
• Used train/test split and model validation techniques for robust evaluation

---

📂 Project Structure

|- Crime Data Project
    |- Data Cleaning & Preprocessing (Task 1)
        |- Handle missing data  
        |- Fix data types  
    |- Feature Engineering (Task 1)
        |- Create time-based features  
        |- Aggregate crime counts  
    |- Visualization (Task 1)
        |- Bar charts  
        |- Heatmaps  
        |- Line charts  
    |- ML Pipeline (Task 2)
        |- Data preprocessing  
        |- Model selection and training  
        |- Model evaluation and testing  
    |- Reports (Tasks 1 & 2)
        |- ydata_profiling HTML report  
        |- Jupyter notebooks  

---

📁 File Structure

Task 1:

• Cleaned Crime Data Set.csv – Raw crime dataset
• CrimeData.ipynb – Jupyter Notebook for cleaning, analysis, and visualization
• CrimeData.html – Exported interactive HTML report
• main.py – Python script for preprocessing, feature engineering, and plotting

Task 2:

• Crime Data Analysis.ipynb - Scikit-learn Pipline
• data.csv - Raw Crime dataset

---

🛠️ Installation and Setup

Clone the repository and install required dependencies:

git clone https://github.com/ViggiXDevp/DataScience_MachineLearning.git  
cd DataScience_MachineLearning  
pip install pandas ydata_profiling  

---

▶️ Running the Project

1. Open CrimeData.ipynb in Jupyter Notebook or Google Colab to run Task 1 for EDA and visualizations.
2. Execute the notebook cells sequentially to load data, clean, analyze, and generate visualizations.
3. Run main.py for scripted preprocessing, feature engineering, and plotting (Task 1).
4. Open and run Crime Data Analysis.ipynb for Task 2 to train, test, and evaluate the Scikit-learn pipeline.
5. Review model performance and insights generated from the machine learning pipeline.
6. View the interactive report CrimeData.html for exploratory analysis insights.

---

🌐 Google Colab Link

You can also run this project interactively in Google Colab without any local setup:

Crime Data EDA & Visualisation

---

💡 Challenges & Creative Additions

• Handled missing values and cleaned data for reliable analysis.
• Created feature engineering techniques to extract temporal trends.
• Used ydata_profiling to generate a detailed automated profiling report.
• Modularized scripts for ease of scalability and reusability.
• Developed and fine-tuned a Scikit-learn pipeline for crime prediction.
• Incorporated appropriate model selection and evaluation strategies for robust results

---

🎓 Submission Details

• GitHub Repository: https://github.com/ViggiXDevp/DataScience_MachineLearning
• Full Name: G V VIGHNESH REDDY
• Email: gvvighneshreddy8612@gmail.com
• Registration Number: 24BAI10374

---

⚖️ License

This project is licensed under the MIT License.

---

🧑‍💻 Author & Contact

ViggiXDevp
✉️ Email: gvvighneshreddy8612@gmail.com
🐙 GitHub: https://github.com/ViggiXDevp
🔗 LinkedIn: https://www.linkedin.com/in/vighneshreddy