This will predict whether any car is worth buying. It will produce four classes Unacceptable, acceptable, good, very good.
We have used multiclass neural network model to train the data. Database is obtained from https://archive.ics.uci.edu/ml/machine-learning-databases/car/car.data. This database has eight attributes and one class attributes. Class attributes include Unacceptable, acceptable, good, very good. We have tried other models like decision tree, forest, jungle for classification but the maximum accuracy gained through any of this is around 85%. Using multiclass neural network accuracy achieved by us was 95% at first but after changing the learning rate and number of hidden layers we have increased our overall accuracy to 98%.