Page 176 - Robotics and AI class 10
P. 176
Regression is a mathematical approach to find a relationship between two or more variables. It works with
continuous data. This can be used for weather forecasting, time series modelling, etc. In order to get the best fit
results, the distance between the line and data points should be minimum.
Types of Regression:
• Linear Regression: Linear Regression is the most used regression modeling technique. It assumes that there is a
linear relation between a dependent variable and independent variable, which is reflected with a regression line
and is a best fit line. This linear relation is represented as Y=b*X+c+e. “X” and “Y” represents independent and
dependent variables respectively, “b” denotes slope of the line, “c” denotes interception and “e” denotes error
and the linear connection is defined as Y = c+m*X + e, where ‘c’ denotes the intercept, ‘m’ denotes the slope of
the line, and ‘e’ denotes error in the model.
• Multiple Linear Regression: Multiple linear regression is a statistical technique used to model the relationship
between one dependent variable and two or more independent variables. The dependent variable is typically
continuous and measured on an interval or ratio scale. The independent variables can also be continuous and
measured on an interval or ratio scale.
• Logistic Regression: If the dependent variable is discrete, that is it has only 2 values like 0 or 1, true or false,
pass or fail, etc, the logistic regression technique can be used. In this the target variable can only have one of two
values, and a sigmoid curve represents connection between the target and the independent variables.
Few things have to be checked in the data before Logistic Regression is used, that the data should not have
multicollinearity and the data is large with approximately equal occurrences of value to come in target variables.
When the dependent variable only has 2 values of 0 or 1, it is referred to as Binary or Binomial Logistic Regression.
Classification
This is another rule-based AI model. It is a systematic grouping of observations in categories, something like
categorising plants, animals in different taxonomies by biologists. In classification you teach the machine to
perform with labelled data. Testing data is then classified as one of the labels of the training dataset. The algorithm
is able to determine which set a given data point belongs to by means of a classification function represented by
the dotted line. The model classifies datasets according to the rules given to it.
For example, if we want to train a model to identify if an image is of an onion or a ginger, we need to train it with
multiple images of both onion and ginger along with their labels. The machine will then classify images on the
basis of the labels and predict the correct label for testing data. Classification works on discrete dataset.
174 Touchpad Robotics & Artificial Intelligence-X

