In this video we introduce how to define and train a classification machine learning model using matlab's neural network toolbox, and discuss network complexity and over training, as well as how to analyze the neural network with KPIs, including the receiver operator (ROC) curve, a confusion matrix, and cross-entropy.
For the example in the video we classify color data from Randall Munroe. His blog post with the full raw data set is available here:
blog.xkcd.com/2010/05/03/colo...
If you want to follow along with the analysis, a short version of the dataset and the example code can be found here:
github.com/codinglikemad/Matl...
My previous video on neural nets in matlab can be found here:
• Neural Networks in Mat...
Music Credits:
The music in the final section of the video is "Ticker" by "Silent Partner".
29 июн 2024