Our code is hosted at http://server.einsteinly.me

The log in token for our Jupyter server is einsteinly

Our whole programme

import tensorflow as tf
import keras
import cv2
import numpy as np
import matplotlib.pyplot as plt
import os
import sklearn
import pandas as pd
from sklearn.model_selection import train_test_split
import pickle
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D

# config params
batch_size = 50
epochs = 30

train = pickle.load(open('./data.pickle', 'rb'))
X_train = train['X_train']
X_test = train['X_test']
y_train = train['y_train']
y_test = train['y_test']

model = Sequential()
model.add(Conv2D(4, (3, 3), activation='relu', input_shape=(100, 100, 1)))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(8, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(16, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(32, (3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(Dropout(0.4))


model.add(Flatten())
model.add(Dense(32, activation='softmax'))
model.add(Dense(5, activation='softmax'))

model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])

model.fit(X_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(X_test, y_test), shuffle=True)
results = model.evaluate(X_test, y_test, batch_size=batch_size)
print(results[1])
model.save_weights("./output/weights_" + str(results[1]) + ".convnet")



file = open("test.output","w")
print("testing model")

# loading test set
test = pickle.load(open('./test.pickle', 'rb'))
X_test_raw = test["X_test_raw"]
X_test_raw = np.array(X_test_raw)
X_test_raw = np.resize(X_test_raw, (X_test_raw.shape[0], X_test_raw.shape[1], X_test_raw.shape[2], 1))   


types = ['A', 'L', 'R', 'T', 'W']
results = model.predict(X_test_raw)
#print(results)

for i in range(len(results)):
    print("Prediction for fingerprint " + str(i + 1) + " : " + types[np.argmax(results[i])])
    file.write("Prediction for fingerprint " + str(i + 1) + " : " + types[np.argmax(results[i])] + "\n")
    for j in range(len(results[i])):
        print("   Probability that it is "+ types[j] + " : " + str(results[i][j]))
        file.write("   Probability that it is "+ types[j] + " : " + str(results[i][j]))

file.close()

Built With

• python