Add files via upload

Author: harshraj3223

Conclusion.txt

@@ -0,0 +1,9 @@
+In the paper, a total of seven CNNs are trained. The first CNN for predicting the square from where a move has to be made. The other six CNNs
+are for the prediction for each of the six distinct pieces in a chess (considering white's turn) to go to a particular position as a move.
+
+In our project, we have considered training only two CNNs, which'll perfectly suffice. The first CNN for predicting the square from where
+a move has to be made. The second CNN for predicting the square to where a move has to be made ( given the board layout ).
+
+In Paper: test/validation accuracies of CNNs: {38.3%}, {52.20%, 29.25%, 56.15%, 40.54%, 26.52%, 47.29%}.
+
+In our project: test/validation accuracies of CNNs: {48.65%}, {27.23%}