This video is the second and final part of our series on Cost Functions in Artificial Intelligence.
In this tutorial, Mr. Bharani Kumar Depuru, Director of 360digiTMG teaches us how to evaluate the performance of our Machine Learning Algorithms and Deep Learning Algorithms with the help of cost functions. This video is a must-watch for data scientists and IT professionals.
Whenever we use Deep Learning Algorithms or Machine Learning Algorithms to build an Artificial Intelligence Model there is the probability of generating errors. We need to measure the performance of a Machine Learning model by validation and testing. We need to calculate the error which is the difference between the predicted values and expected values of output variables and present it in the form of a single real number. Cost functions or error functions help us calculate and minimize the errors in our Machine Learning Model.
This video tutorial deals with entropy, cross-entropy and information gain. These are cost functions used in classification problems in machine learning algorithms. Entropy is a measure of disorder or uncertainty and the goal of machine learning models and data scientists is to reduce entropy. Entropy is treated as an error in machine learning algorithms. The entropy equation is derived and the concept of information gain is explained with the help of a lucid example. In any model, it is wisest to choose those variables or outcomes that give you the highest information gain. In such a situation of the highest information gain entropy or error is zero. This video teaches the student how to model a problem to attain zero entropy.
The important topic of cross-entropy is discussed with example. Cross entropy is the variance or distance between two probability distributions. Cross entropy is used for evaluating training data and show how far the predicted probability distribution is from the actual one. The two types of cross-entropy Categorical Cross-Entropy and Binary Cross Entropy are discussed with examples and equations are derived for the same.
This is a classic video on the theoretical aspects of entropy, information gain and cross-entropy and the case examples are thorough and intelligent. All students of data science will greatly benefit from this tutorial.
SUBSCRIBE TO 360DigiTMG’s YOUTUBE CHANNEL NOW
https://www.youtube.com/channel/UCNGIDQ466bNY87eEeKeQuzA
We have specifically created a Facebook Group for all our Data Science aspirants. You can use the below link to join.
In addition to this, we are going to host 2 FREE training sessions Every Single Month on various topics inside this group.
Join FREE Data Science Facebook Group
https://www.facebook.com/groups/DataScience.MachineLearning.ArtificialIntellegence/
CONNECT WITH 360DigiTMG ON SOCIAL MEDIA
Facebook: https://www.facebook.com/360Digitmg/
Linkedin: https://www.linkedin.com/company/360digitmg/
Instagram: https://www.instagram.com/360digitmgindia/
YouTube: https://www.youtube.com/channel/UCNGIDQ466bNY87eEeKeQuzA
About 360DigiTMG
360digiTMG is a 5-year-old training & consulting organization led by stalwarts of the industry who are alumnus of premier institutions like the Indian Institute of Technology, Indian Institute of Management and Indian School of Business. 360digiTMG since its inception has been the forerunner in the space of management and niche programs that aid in up-skilling and cross skilling executives across various levels and domains. 360digiTMG has been conducting training programs across the globe for corporate and individuals alike.
360DigiTMG is one stop solution to all the trainings in emerging technologies such as Artificial Intelligence, Machine Learning, Big Data, Project Management, Quality Management, etc. 360DigiTMG is a training company, which is a division of the analytics consulting firm Innodatatics Inc.
For more Information Contact us @::
India : +91 99899 94319
Malaysia: +603 2092 9488
Email: [email protected]
Web: https://360digitmg.com/
Did you find this video helpful? Leave a comment below!
#DataScience #ArtificialIntelligence #Scholarship #DataAnalytics #Jumpstart #360DigiTMG #Malaysia