Learning Goal: I’m working on a artificial intelligence project and need support

Learning Goal: I’m working on a artificial intelligence project and need support to help me learn.THE DATASET The exfiltration dataset uses 20s histograms of the 3-gram of system calls to predict whether a process is malign (exfiltrating) or benign. The dataset consist of 1942 input features/attributes which are the system call 3-grams and 2 classes/labels with column name ‘Label’ in the csv file (0 for benign and 1 for malign) . There are a total of 39,264 samples in the dataset (23,134 benign and 16,130 malign). All variable types are integers and correspond to the raw number of occurrence of each sequence in a 20s time period.The assignment consists of training a Generative Adversarial Network (GAN) for the provided dataset. Choose and train a baseline model for the dataset classification
Train a discriminator/generator pair on the dataset
Use the output of the GAN to train a classifier
Compare the classification accuracy to the baseline
A tutorial on GAN training using the Keras dataset can be found here: https://machinelearningmastery.com/how-to-code-the-generative-adversarial-network-training-algorithm-and-loss-functions/ The classification performance is most likely slightly worse when the discriminator is trained with a generator. There are papers that focus more on improving classification accuracy with GANs as opposed to generating realistic looking samples. It seems that it is possible to increase performance but the samples look worse with these techniques. It just depends on what you are trying to achieve. I need help to convert the dataset into images (heatmaps or histograms of some sort), then do the training/classification & GAN on the images instead of doing on the original dataset as csv.Dataset link: https://www.dropbox.com/s/yqxrh0whalgjudh/Labelled_histograms.csv?dl=0
Requirements: 10 slides + python code