Tutor profile: Mdredwan I.
Subject: Python Programming
Suppose you want to build a website where you want to upload an image and get its classification using only Python. What are the libraries you will need for that?
Here, Python is being used for Machine Learning and Deep Learning, Data Processing and Website Building. For the website building, we can use either Python's Flask or Django Framework. For Data Processing we have to use NumPy, SciPy and Pandas. For image reading and classification we may use only PyTorch or also use OpenCV-Python as well for easy reading.
What are the toolboxes that allow you to do Image Classification in MATLAB?
It depends of the approach of image classification process. if we follow the process of manual image classification, i.e. extract features manually, then we can use the Signal Processing Toolbax and Image Processing Toolbox to manually extract the features and then use the Statistics and Machine Learning Toolbox to do the classification. Or if we want to use Deep Learning, there is a Deep Learning Toolbox which contains all the necessary Deep learning architectures such as convolution2dLayer, imageInputLayer, resnet101 , etc for doing the job.
Subject: Machine Learning
In a Decision Tree, how can we select the features to make the branches of the tree?
A good feature is a feature that lets us make correct classification decision or at least the most correct decision for most of the samples. Some of the criteria to measure how good a feature is to branch out the decision tree is 1. Classification accuracy 2. Information Gain (using entropy) 3. The GINI Index (used in CART, IBM IntelligentMiner)