CSCI-561 - Fall 2023 - Foundations of Artificial Intelligence
Homework 3
Due Time: Monday, November 20th, 2023, 23:59:59 PST
Overview
This homework explores the applications of Temporal Reasoning in Artificial Intelligence. In general, the
solution for a temporal reasoning task involves taking a sequence of actions/observations on an Partially
Observable Markov Decision Process (POMDP Environment) , applying a temporal-reasoning algorithm
that you learned from this class, and returning the most probable sequence of the hidden states that the
POMDP most-likely went through when experiencing the given sequence of actions/observations.
More specifically, this assignment provides you with two versions of temporal data: a base version
involving the “Little Prince” Environment and an advanced version that revolves around speech
recognition and text prediction.
Scenario 1 : Little Prince Model
The setup in this version is very similar to the “Little Prince” environment shown in Figure 1, presented in
the lecture notes and in the optional reading textbook (ALFE).
Figure 1: The Little Prince POMDP (lecture 08-09 and 22).
You will be given a list of available percepts, actions and states and the corresponding initial state
weightages, transition and observation weight values in that environment (More on the input structure will
be covered in the sections below) . Your task is to design and implement a temporal-reasoning algorithm,
that will take a sequence of actions/observations and determine the most-likely sequence of states that this
POMDP has gone through, as shown in Figure 2. For example, if the Little Prince’s experience is given as