You are on page 1of 3

COM 418: Expert Systems

NAME : MBURU KENNEDY


REG NO : COM/0808/20
Due : Tuesday 1st August 2023 :

1. Describe how uncertainty occurs in the knowledge domain (2 marks)


From weak implications/vague associations- Available information in the domain may be
incomplete or imperfect making it difficult to have concrete associations between antecedent and
consequent.
From imprecise language or vague terms which can make it difficult to express knowledge in
precise IF...THEN... form of production rules.
Incomplete data or unknown data - missing data or gap in knowledge base can make it
challenging reaching definitive conclusions.
Conflicting expert opinions - contradictory expert opinions can produce conflicting rules.
2. Discuss the A* search technique ( 2marks)

A* search is an informed algorithm used in expert systems for finding the shortest path or optimal
solution by combining the cost to reach a node from the start (g-value) and the estimated cost to
reach the goal from that node (h-value), efficiently guiding the search towards the best path.
3. Explain the importance of reinforcement learning algorithms (2 marks)

Reinforcement learning is a type of machine learning that allows agents to learn how to behave in
an environment by trial and error. Agents are rewarded for taking actions that lead to desired
outcomes, and penalized for taking actions that lead to undesired outcomes. Over time, agents
learn to take actions that maximize their rewards.

 They can learn from experience: Reinforcement learning algorithms can learn from their
interactions with the environment, without being explicitly programmed. This makes them well-
suited for solving problems in real-world environments, where the rules are not always known.

 They can adapt to change: Reinforcement learning algorithms can adapt to changes in the
environment. This makes them robust to changes in the environment, and allows them to
continue to perform well even when the environment changes.

4. Discuss mamdani inferencing as used in Fuzzy logics. (3 marks

Mamdani inference is a method in fuzzy logic for converting linguistic rules into precise numerical
outputs.
It involves three main steps: fuzzification, rule evaluation, and defuzzification.
In the fuzzification stage, linguistic variables are converted into fuzzy sets using membership
functions.
Then, fuzzy rules are evaluated based on the degrees of membership of the input variables.
Aggregation methods combine the outputs of multiple rules.

Defuzzification converts the aggregated fuzzy output into a crisp value.


Mamdani inference allows for handling imprecise or uncertain information, making it valuable for
complex decision-making and control systems in various applications.
5. Discuss k-nearest neighbour (K-NN) Algorithm ( 2 marks)
The k-nearest neighbor (K-NN) algorithm is a non-parametric classification and regression
technique in machine learning.In K-NN classification, when given a new data point, the algorithm
identifies the k nearest data points (neighbors) in the training dataset based on a chosen distance
metric . The majority class among these k neighbors is assigned to the new data point as its
predicted class.
6. Briefly explain what Machine learning entails (3 marks)

Machine learning is a subset of artificial intelligence (AI) that involves the development of
algorithms and models that enable computers to learn from data and improve their performance
on specific tasks without being explicitly programmed.
It is a data-driven approach where machines use patterns and statistical techniques to make
predictions, recognize patterns, classify objects, or perform other tasks.

There are three main types of machine learning:

Supervised learning: This type of machine learning involves using labeled datasets to train


algorithms for accurate classification or outcome prediction.

Unsupervised learning: This type of machine learning involves algorithms that train on unlabeled
data. The algorithm scans through data sets looking for any meaningful connection.
Reinforcement Learning: Reinforcement learning involves an agent learning by interacting with an
environment and receiving feedback in the form of rewards or penalties. The goal is to find the
optimal policy to maximize cumulative rewards over time.

7. Explain the functions of the following parts of an expert system. (2 marks)


i. Knowledgebase
Stores domain-specific knowledge in a structured manner. It contains information about rules,
facts, and relationships relevant to the domain the expert system operates in.
ii. Database of facts
Stores the current set of facts or data that the expert system uses during its operation. Provides a
place to store and manage the data received from users or external sources.
8. What is the role of the membership sets in fuzzy logics (2 Marks)
To represent the degree of membership or the degree of truth of an element to a particular fuzzy
set.
Providing a mechanism to model and reason with imprecise and uncertain
information

9. Name TWO factors that make fuzzy logic a preferred technique over expert systems. ( 2
Marks)

Fuzzy logic can handle uncertainty: Expert systems are based on the concept of "if-then"
rules,which can be difficult to apply to problems with uncertainty. Fuzzy logic, on the other hand,
is able to handle uncertainty by using membership functions to represent the degree to which an
element belongs to a set.
Linguistic Modeling: Enables experts to describe complex relationships and rules in natural
language terms. This makes it easier for domain experts to express their knowledge without the
need for formal, mathematical representations

You might also like