Page 14 - Artificial Intellegence_v2.0_Class_11
P. 14
Correlation and Regression Knowledge – Correlations, Regression, and other related terms
o Crosstabs and scatterplots Applications – Being able to relate data with regression and
o Pearson's r correlation. Everyday applications of these mathematical
o Regression - Finding the line concepts.
o Regression - Describing the line
o Regression - How good is the line?
REGRESSION o Example contingency table
Correlation is not causation
o
o
Example Pearson's r and regression Readings
o o Correlation
Regression
o Caveats and examples
o Practice exercise Correlation and Regression
o Explain the importance of data from above examples
o How prediction changes with changing data?
What is a classification problem? Knowledge – What is classification and its types, what kind of
Examples problems may be placed under the category of a classification
- Simple binary classification
Introduction to binary classification with logistic regression problem
CLASSIFICATION&CLUSTERING Practice exercise on simple Binary Classification model society
Applications – Where to apply classification principals
Analysis – Impact of the application of incorrect algorithms on
True positives, true negatives, false positives and false
negatives
Knowledge – Clustering problems and its application, why is it
Where we should care more with examples
o
called clustering
o
Example- false negative of a disease detection can have
different implication than false positive, one will be
Application – Application of clustering problem using standard
more physical harm and other will be mental
models
What is a clustering problem?
Why is it unsupervised?
Examples
Practice exercise on simple Clustering model Knowledge – What is ethics, Impact of ethics on society, the
AI working for good
AI VALUES (BIAS AWARENESS) Principles for ethical AI impact of bias on AI functioning
Evaluation – Biases in data, how to de-bias or neutralize the
Types of bias (personal /cultural /societal)
biased data
How bias influences AI based decisions
Application – Finding bias in acquired dataset
How data driven decisions can be debiased
Hands on exercise to Detect the Bias (Intro to AI)
(xii)

