WEKA Tutorial #1.2 - How to Build a Data Mining Model from Scratch

[@yasmirathnayaka8635]

Can all 3 options (Use training set, cross-validation & percentage split) available here be used to train the model? Isn't cross-validation an evaluation method? Can you explain these 3 options separately? and will you tell me how to save the trained model, please?

[@jeffersonjones7863]

0:04 quite humid in Bangkok 🤔🌁 Thanks for further instructions though! 🙏🏼

[@DataProfessor]

Yes, it is! Rainy weather at the moment.

[@abhipsatripathy3934]

Please explain no. of leaves and size of the tree....what does that actually mean? And also explain the matrix just above the confusion matrix.

[@DataProfessor]

In a nutshell, the leafs are the terminal nodes where decisions are made while the branches are associated with a decision threshold whereby particular values are used as threshold to dictate the path for subsequent branching nodes. Collectively, all branches will lead to a series of if-then rules that culminates in the final decisions of a prediction of which class label each samples are classified as. An introductory explanation of decision trees is provided by this medium article, link.medium.com/czCUVrNfV7. Results interpretation is provided by this post, stackoverflow.com/questions/2903933/how-to-interpret-weka-classification

[@abhipsatripathy3934]

From where can I download iris data set? can you provide me a link? Sir, for WEKA to use, we need data file in arff extension, where to get that?

[@DataProfessor]

Hi, you can download the arff files from this website and try to see the files in a text editor to see how to create for a new dataset storm.cis.fordham.edu/~gweiss/data-mining/datasets.html

[@abhipsatripathy3934]

@@DataProfessor Thank you so much Prof.

[@sembutininverse]

❤️🙏🏻❤️

[@DataProfessor]

Thanks for watching!