Trjtdtk

I'm trying to train a machine learning model on the following data.
Here is a sample:

key A0 A1 A2 A3 A4 A5 A6 A7 A8 A9

a 12 15 28 35 22 10 11 10 18 11

d 11 14 29 24 10 11 10 19 11

The model is supposed to classify every letter of the alphabet and get
a non-random row of numbers associated with the letter. A-Z, with
values ranging from about 8-35. I've had to resort to using tensorflow 1.5.0 because I don't have an expensive GPU to train on.

If I run for only 10 epochs or less I get reasonable predictions of
about .04 percent, which would be pretty close to a random guess. But
if I try to train for longer I get the following.

[4.8728631e-04 2.8466644e-02 8.8677540e-02 ... 6.2057756e-02
2.9221788e-02 8.6455815e-02]
[1.4348865e-01 1.2017406e-01 3.5096225e-03 ... 3.7368879e-02
5.3554219e-03 1.1316765e-03]
[2.2863398e-06 1.4322808e-03 5.2052658e-02 ... 6.3502453e-03

This is the model that isn't working
import pandas as pd
import tensorflow
import numpy as np
import keras
from keras.models import model_from_json
dataset = pd.read_csv('mlglovesdata.csv')
X = dataset.iloc[:, 1:11].values
y = dataset.iloc[:, 0].values

# Taking care of missing data
from sklearn.preprocessing import Imputer
imputer = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0)
imputer = imputer.fit(X[:, 1:11])
X[:, 1:11] = imputer.transform(X[:, 1:11])

# dataset starts at 1 due to columns from openoffice formatting, I suppose
# Encoding categorical data
# Encoding the Dependent Variable
from sklearn.preprocessing import LabelEncoder, OneHotEncoder
labelencoder_y = LabelEncoder()
y = labelencoder_y.fit_transform(y)
# May need to only go this far in the encoding
onehotencoder = OneHotEncoder(sparse=False)
y = y.reshape(len(y), 1)
y = onehotencoder.fit_transform(y)
y = y[:, 1:]
# Splitting the dataset into the Training set and Test set
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)


# Feature Scaling
from sklearn.preprocessing import StandardScaler
sc_X = StandardScaler()
X_train = sc_X.fit_transform(X_train)
X_test = sc_X.transform(X_test)
# Importin keras dependencies
import keras
from keras.models import Sequential
from keras.layers import Dense
# Initializing the ANN
classifier = Sequential()

# Add 1st input hidden layer
classifier.add(Dense(17,kernel_initializer='uniform',activation='relu',input_dim = 10))

# Add second hidden layer
classifier.add(Dense(17,kernel_initializer='uniform',activation='relu'))

# Add the output layer
classifier.add(Dense(25,kernel_initializer='uniform',activation='softmax'))



classifier.compile(loss="categorical_crossentropy", optimizer="adam", metrics=['accuracy'])


classifier.fit(X_train,y_train,batch_size=10,nb_epoch=20)


y_pred = classifier.predict(X_test)
print(y_pred)

Any help would be a greatly appreciated, and might earn an honorable
mention in a book if I write one.