Filter by:
๐Ÿ“ˆ
Machine Learning

Linear Regression Model

Build your first ML model from scratch. Implement linear regression to predict house prices using gradient descent and evaluate model performance with metrics.

code Python
functions Mathematics
Difficulty:
schedule 4-6 hours
๐Ÿ–ผ๏ธ
Computer Vision

CNN Image Classifier

Design and train a Convolutional Neural Network using TensorFlow or PyTorch. Classify images from CIFAR-10 dataset and implement data augmentation techniques.

layers Deep Learning
image CNN
Difficulty:
schedule 8-10 hours
๐Ÿ’ฌ
๐Ÿ˜Š
๐Ÿ˜
๐Ÿ˜ข
NLP

Sentiment Analysis System

Build an NLP model to analyze sentiment in product reviews. Use transformers or LSTM networks to classify text as positive, negative, or neutral sentiment.

text_fields Text Analysis
psychology Transformers
Difficulty:
schedule 6-8 hours
๐ŸŽฏ
Recommendation

Movie Recommender System

Create a collaborative filtering system for movie recommendations. Implement both user-based and item-based filtering using the MovieLens dataset.

explore Filtering
stars Ratings
Difficulty:
schedule 10-12 hours
๐ŸŽฅ
Computer Vision

Real-time Object Detection

Implement YOLO or Faster R-CNN for real-time object detection. Process video streams and identify multiple objects with bounding boxes and confidence scores.

videocam Real-time
speed YOLO
Difficulty:
schedule 12-15 hours
๐Ÿค–
NLP

Conversational AI Chatbot

Build an intelligent chatbot using transformer models. Implement context awareness, intent recognition, and generate human-like responses using GPT or BERT.

chat_bubble Dialogue
psychology_alt LLM
Difficulty:
schedule 15-20 hours
๐Ÿ“Š
Time Series

Stock Price Forecasting

Predict stock prices using LSTM or GRU networks. Handle time series data, implement feature engineering, and evaluate forecasting accuracy with RMSE.

trending_up Forecasting
show_chart LSTM
Difficulty:
schedule 8-10 hours
๐ŸŽจ
Generative AI

Generate Images with GANs

Create synthetic images using Generative Adversarial Networks. Train generator and discriminator networks to produce realistic faces or artwork.

auto_awesome Generation
brush Creative
Difficulty:
schedule 15-18 hours
๐Ÿ”ต
Unsupervised

Customer Segmentation

Segment customers using K-Means clustering. Analyze customer behavior patterns, visualize clusters, and derive actionable business insights from data.

group_work Clustering
insights Analytics
Difficulty:
schedule 5-7 hours
๐ŸŽฎ
RL

Train Gaming AI Agent

Use Q-Learning or Deep Q-Networks to train an agent to play games. Implement reward systems, explore-exploit strategies, and optimize agent performance.

psychology RL
sports_esports Gaming
Difficulty:
schedule 12-15 hours
โš–๏ธ
๐Ÿ“„
๐Ÿ”
โœ๏ธ
Ethics

AI Ethics & Bias Analysis

Research ethical implications of AI systems. Analyze bias in AI models, explore fairness metrics, and propose solutions for responsible AI development.

balance Ethics
article Research
Difficulty:
schedule 6-8 hours
๐Ÿง 
Deep Learning

Build Neural Network from Scratch

Implement a feedforward neural network without using libraries. Code forward propagation, backpropagation, and gradient descent from first principles.

memory Fundamentals
calculate Math
Difficulty:
schedule 10-12 hours