ML-PROJECTS

Machine Learning Internship – Task 1 to 4 Submission

Intern: Savanth G

B.Tech CSE (AIML)
Presidency University, Bengaluru
Internship Period: 9 June 2025 - 9 july 2025

---

Overview

This repository contains the completed Task 1, Task 2, Task 3, and Task 4 of the Machine Learning Internship. These tasks involve building intelligent ML systems for restaurant data: including rating prediction, recommendation engines, cuisine classification, and sentiment analysis. Each task demonstrates hands-on use of data preprocessing, machine learning models, NLP, and performance evaluation.

Task Summary

Task 1: Restaurant Rating Prediction

• Objective: Predict the aggregate rating of restaurants using various features such as price range, votes, and location.
• Steps Covered:
  • Data cleaning and missing value handling
  • Encoding categorical features using get_dummies
  • Model training with:
    • RandomForestRegressor
    • LinearRegression
  • Model evaluation using:
    • R² Score
    • Mean Squared Error (MSE)
  • Feature importance analysis for interpretability

---

Task 2: Restaurant Recommendation System

• Objective: Build a content-based filtering system that recommends restaurants based on user preferences.
• Features Used:
  • Cuisine type
  • Price range
  • Rating threshold
  • Delivery availability
  • City and currency filters
• Workflow:
  • Label encoding of user filters
  • Text vectorization using TfidfVectorizer
  • Similarity matching using CosineSimilarity
  • Dynamic user input handled with validations and defaults
• Output: A sorted table of the top 10 recommended restaurants matching user-defined criteria.

---

Task 3: Cuisine Classification

• Objective: Classify restaurants into cuisine categories using supervised classification algorithms.
• Steps Covered:
  • Data preprocessing: handling missing values & encoding categorical variables
  • Train-test split
  • Model training with:
    • RandomForestRegressor
    • LinearRegression
• Evaluation using:
  • Accuracy
  • Precision
  • Recall
  • Addressed class imbalance and analyzed model performance across cuisine categories

---

Task 4: Sentiment Analysis on Restaurant Reviews

• Objective: Analyze the sentiment (positive/negative) of customer reviews using Natural Language Processing (NLP).
• Steps Covered:
  • Text cleaning: punctuation removal, stopword removal, lowercase conversion
  • Tokenization and lemmatization
  • Feature extraction using TfidfVectorizer
  • Model training with:
    • LogisticRegression
    • MultinomialNB
• Evaluation using:
  • Accuracy
  • Confusion Matrix
  • Classification Report
• Output: Classifies reviews as Positive or Negative. Includes sentiment prediction for both batch data and custom inputs.

---

Tools & Libraries Used

• Python, pandas, numpy
• scikit-learn, NLTK/spaCy
• TfidfVectorizer, LabelEncoder
• Google Colab
• ML Algorithms

---

File Structure

ML-PROJECTS/

• Task1_Rating_Prediction.ipynb
• Task2_Recommendation_System.ipynb
• Task3_Cuisine_Classification.ipynb
• Task4_Sentiment_Analysis.ipynb
• Dataset.csv
• README.md

Contact

Savanth G
📧 savanthg14@gmail.com
🌐 LinkedIn : linkedin.com/in/savanth-g-65454a36b