Artificial Intelligence Nanodegree Projects

1. Sudoku

• Constraint propagation to solve a sudoku (diagonal and naked twins)

2. Game Agent

• Minimax algorithm is implemented
• Minimax algorithm with alpha-beta pruning with iterative deepening is implemented
• 3 custom heuristics are implemented and compared
• Research review of Game Tree Searching by Min/Max Approximation is written
• Sketchnotes MiniMax and AlphaBeta Pruning

3. Adversarial Planning Search

• The 3 given air cargo problems are implemented
• Uninformed planning algorthims implemented: -- Breath-first-search -- Depth-first-search -- Uniform-cost-search
• A-Star search is implemented using automatic heuristics
• provide optiomal sequence for each problem
• Sketchnotes Solving Search Problems with AI

4. Sign Language Recognition

• Model selection techniques BIC, DIC and K-fold Cross Validation are used
• Trained a set of Hidden Markov Models (HMMs) using part of a preprocessed dataset of tracked hand and nose positions extracted from video to try and identify individual words from test sequences.
• Sketchnotes for HMMs

5. Dog Breed Classifier

• Built an algorithm to identify canine breed given an image of a dog. If given image of a human, the algorithm identifies a resembling dog breed.
• Sketchnotes for Neural Networks

6. Recurrent Neural Networks

• Built RNNs that can generate sequences based on input data
• with a focus on two applications: used real market data in order to predict future Apple stock prices using an RNN model.

7. Facial Keypoint Detector

• Trained a Convolutional Neural Network with custom architecture
• Implemented image processing like De-noise, edge detection and blur

Certificate

Resources:

• Udacity AIND
• Artifical intelligence, a modern approach