sign-language-prediction

Machine learning models used to predict signs for the deaf.

• The project data(signs) has been collected using Microsoft Xbox Kinect.
• The data was then saved to an Microsoft Excel file.

Using machine learning to predict signs.

The data was divided into 75-25 ratio for train and test data. Then the following machine learning algorithms were used to build models:

• Support Vector Machine(SVM)
• k-Nearest Neighbor(kNN)
• Random Forest(RF) Then all of them were used to predict values on the test set.

Then I combined all these predictors and ensembled them to improve accuracy. The overall accuracy of the model was improved.

I plotted the accuracy tables where you can see accuracy of each algorithm for each possible outcome label. I also have plotted graphs for this table for better visualization as to which algorithms predicts better. And then I finally plotted graph for the ensembled model as well.