In this project we where given A task that consists of four notebooks.
-
In the First notebook we aimed to improve our classification notebook from last semester using our new techniques and methods which we learnt this semester.
-
In the second notebook we recieved the "Fashion Mnist" data set and aimed to create the best model, using A minimul amount of featurs, which can predict what image fits to what label. For this notebook we also dealt with A seperete test set for which we did not touch until the end of the notebook.
-
In the third notebook we recieved the Cats vs Dogs data set. This data set consists of many pictures of dogs and cats. Our goal was to create the ultimate model to predict if A certain given picture is A dog or A cat
-
In the last notebook we did not recieve A data set. Instead we recieved many recordings of hand movevents in 3 different states : spontany,alone and sync. We had to load and prepare the data and turn it into A proper dataset. Then we cleaned , visualizad and continued just as we have in the previous notebooks.
Classifications Models and methods we used : KNN , Random Forest, XGBoost, AdaBoost , Voting(soft and hard) , Bagging, Pasting, Stacking, K-Means , Mini Batches, Logistic Regression, Gaussian Naive Bayes.
Additional information about the techniques and models I used, can be found in the pdf in this repository.
