SUTD 50.021 Artificial Intelligence Coursework. For more information, refer to https://istd.sutd.edu.sg/undergraduate/courses/50021-artificial-intelligence.
Solving the Word Ladder & Flight Itinerary problems using tree search (breadth_first_tree_search, depth_first_tree_search) and graph search (depth_first_graph_search, breadth_first_graph_search, uniform_cost_search, depth_limited_search, iterative_deepening_search and astar_search).
Word Ladder: In a word ladder puzzle, you are given a start word and a goal word. You must transform the start word into the goal word by transforming one letter at a time,
and all intermediate words must be valid as well.
Flight Itinerary: A flight has a starting city and time and an ending city and time. Cities are represented as strings and times as integers. Using a start city, starting time, destination city and deadline time, find the sequence of flights, starting in the starting city, leaving sometime at or after the starting time, and arriving at the destination before (or at) the deadline.
Consider a 3x3 array where each of whose entries can be either 1, 2 or 3. Find an assignment to each of the entries so that the entries in each row, in each column and in one of the diagonals are different.
Input domain/problem files into any online PDDL editor/solver.
Conditional Variational Autoencoder (CVAE) implementation with 5 Conv2d layers.
Pytorch backward hooks for analysing gradient statistics and guided backpropagation to visualise neurons that act like detectors of particular image features.
| Guided BP Example 1 | Guided BP Example 2 | Guided BP Example 3 |
|---|---|---|
 |
 |
 |