In python, make a predictor for a minesweeper game on a 5x5 field. The board can max go…
- User
In python, make a predictor for a minesweeper game on a 5x5 field. The board can max go from 1 to 25.
You’ve data for the past 30 games played, in a list, where each number is an old mine location. You need to use this data. Your goal is to predict a new game with the data. The game itself got 3 mines in it, and you need to predict 5 safe spots and 3 possible mine locations.
You have to use machine learning for this. You need to Deep Learning and make it as fast as possible. Also, another thing, you CAN’T make it random. It cannot be random in anyway and needs to predict out of the machine learning module.
Your data is: [5, 12, 13, 4, 17, 19, 4, 18, 21, 1, 6, 11, 7, 15, 22, 14, 18, 19, 1, 19, 20, 9, 16, 17, 1, 19, 20, 6, 14, 24, 15, 20, 22, 7, 18, 21, 4, 16, 23, 5, 11, 19, 5, 6, 23, 6, 12, 13, 0, 5, 15, 1, 5, 22, 0, 2, 23, 5, 10, 13, 5, 12, 17, 1, 7, 22, 7, 11, 18, 7, 8, 9, 17, 21, 24, 13, 14, 16, 2, 3, 11, 7, 17, 18, 7, 14, 21, 2, 5, 9]
- Assistant
First, let's import the required libraries and preprocess the data.
import numpy as np from keras.models import Sequential from keras.layers import Dense from keras.utils import to_categorical from sklearn.model_selection import train_test_split # data past_games = [5, 12, 13, 4, 17, 19, 4, 18, 21, 1, 6, 11, 7, 15, 22, 14, 18, 19, 1, 19, 20, 9, 16, 17, 1, 19, 20, 6, 14, 24, 15, 20, 22, 7, 18, 21, 4, 16, 23, 5, 11, 19, 5, 6, 23, 6, 12, 13, 0, 5, 15, 1, 5, 22, 0, 2, 23, 5, 10, 13, 5, 12, 17, 1, 7, 22, 7, 11, 18, 7, 8, 9, 17, 21, 24, 13, 14, 16, 2, 3, 11, 7, 17, 18, 7, 14, 21, 2, 5, 9] # preprocess past_games = np.reshape(past_games, (-1, 3)) X = np.arange(1, 26) # train-test split X_train, X_test, y_train, y_test = train_test_split(X, past_games, test_size=0.2, random_state=42)Now, let's build a simple deep learning model using Keras.
# model model = Sequential() model.add(Dense(64, input_dim=1, activation='relu')) model.add(Dense(64, activation='relu')) model.add(Dense(3, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) y_train = to_categorical(y_train, num_classes=26) y_test = to_categorical(y_test, num_classes=26) # training model.fit(X_train, y_train, epochs=10, batch_size=16, validation_data=(X_test, y_test))Finally, let's predict 5 safe spots and 3 possible mine locations.
predictions = model.predict(np.arange(1, 26)) mine_locs = np.argsort(predictions, axis=0)[-3:] safe_spots = np.argsort(predictions, axis=0)[:5] print("Safe spots:", safe_spots.ravel()) print("Mine locations:", mine_locs.ravel())Note that the model above is very small and simplistic. The accuracy might not be very high due to the small dataset, but can be improved by adding more layers or adjusting hyperparameters, or increasing training data.