Downloaded 236 times
Uploaded byKnoldus Inc.
Machine Learning With Logistic Regression
The document discusses machine learning with a focus on logistic regression, including definitions, techniques, and comparisons between regression and classification. It explains logistic regression as a classification algorithm that employs the logit function to evaluate outputs and minimize errors. Key components such as the nature of outputs in regression versus classification and the importance of the logit function in making predictions are highlighted.
In this document
Powered by AI
An introductory overview of machine learning and logistic regression with an agenda outlining key topics.
Machine learning definition, types, and techniques including supervised, unsupervised, and reinforcement learning.
Supervised learning divides into regression and classification, detailing their characteristics and evaluation methods.
Introduction to logistic regression as a classification algorithm using the Logit function for output evaluation.
Description of the logit function, its properties, and the necessity of output value bounding for classification.
List of reference materials for further reading on machine learning and logistic regression.
More Related Content
Similar to Machine Learning With Logistic Regression
More from Knoldus Inc.
Machine Learning With Logistic Regression
- 1. Machine Learning With Logistic Regression AnujSaxena Software Consultant Knoldus Software LLP
- 2. Agenda • Machine Learning– Intro? • Regression vs. Classification • Introduction to Logistic Regression • Learning of Logistic Regression • Use cases
- 3.
- 4. Machine learning ● Machinelearning is the subfield of computer science that gives computers the ability to learn without being programmed.
- 6. Machine Learning techniques ●Supervised ● Unsupervised ● Reinforcement
- 7. Supervised Learning techniques ●There are two categories for supervised learning techniques – Regression – Classification
- 8.
- 10. ● Classification – Outputtype: discrete – Trying to find: a boundary – Evaluation: accuracy ● Regression – Output type: continuous – Trying to find: best fit line – Evaluation: sum of sqaured errors
- 13.
- 14. Logistic Regression ● Atype of classification algorithm ● Based on linear regression to evaluate output and to minimize the error ● Named after the method it uses to evaluate the outputs the Logit function
- 15. What is LogitFunction? ● Odds: Chances of happening one event over chances of this event not happening. ● i.e.
- 17. ● In previousslide we saw the graph of Logit function. ● This Logit function is bounded over x (between 0 and 1 for input values) and unbounded over y-axis (output values) ● But we need y-axis (output values) to be bounded to classify them.
- 18. ● Hence wetake the inverse of the logit function
- 20. References ● Machine Learning– Tom Mitchell ● http://www.theprojectspot.com/tutorial-post