Artificial Intelligence & Machine Learning Task

Instructions:

For the AI & ML Internship, you will need to complete any three of the following tasks for successful completion of the internship. Completing a more advanced task may make you eligible for a Letter of Recommendation.

AI & ML Internship Tasks

Task 1: Sentiment Analysis

Develop a sentiment analysis tool using Python and the Natural Language Toolkit (NLTK). Your task is to analyze text data from social media (such as Twitter or Facebook) to determine whether the sentiment is positive, negative, or neutral. The tool should include data preprocessing steps like tokenization, stop-word removal, and lemmatization.

Task 2: Image Classification

Build an image classification model using TensorFlow or PyTorch. Use a popular dataset like CIFAR-10 or MNIST and create a neural network to classify images into different categories. Implement data augmentation techniques to improve the model's accuracy and visualize the training process using tools like TensorBoard.

Task 3: Predictive Analytics for Sales Forecasting

Create a predictive model for sales forecasting using machine learning algorithms like Linear Regression, Decision Trees, or Random Forests. Use historical sales data to predict future sales. The project should include data cleaning, feature engineering, model training, and validation.

Task 4: Chatbot Development

Design and develop a chatbot using a framework like Rasa or Dialogflow. The chatbot should be able to handle basic customer service inquiries, provide information about products or services, and perform tasks such as booking or ordering. Include natural language understanding (NLU) capabilities to improve the chatbot's responses.

Task 5: Recommendation System

Build a recommendation system using collaborative filtering or content-based filtering techniques. You can choose to work with a dataset like MovieLens or Amazon product data to provide personalized recommendations. The system should be able to handle new user cold-start problems and include evaluation metrics such as precision, recall, and F1-score.