Human are vulnerable to the terrible disease named cancer, which is a major factor in the high mortality rate. There are currently lesser DLTs (Deep learning techniques) or MLTs (machine learning techniques) for identifying cancer, despite advances in cancer treatment approaches. The proposed work performs a comparative study which compares the some significant DLTs like RFs (Random Forests), LSTMs (Long Short Term Memories), CNNs (Convolutional Neural Networks) and BPNNs (Back Propagation Neural Networks). These techniques are used here in this work for classification problem. The techniques are made to classify the medical records into benignand cancerous. Three pathological datasets are used to evaluate the above said techniques. CNNs provide the best performance of 0.97 accuracy and it is even good at its values of precisions, recalls and F1 scores.