How to build a binary classifier with 3D training data

user_8124392 0
Viewed 212

I have data that I have to classify as either 0 or 1. The data is loaded from an .npz file. It gives me training, validation, and test data. This is what they look like:

x_train = [[[  0   0   0 ...   0   1   4]
  [  0   0   0 ...   4  25   2]
  [  6  33  15 ...  33   0   0]
  ...
  [  0  23   4 ...   9  31   0]
  [  4   0   0 ...   0   0  12]
  [  5   0   0 ...   3   0   0]]

 [[ 88  71  59 ...  61  62  62]
  [ 74  88  73 ...  59  70  60]
  [ 69  61  85 ...  60  58  82]
  ...
  [ 68  85  58 ...  55  75  72]
  [ 69  69  70 ...  81  76  83]
  [ 74  68  76 ...  60  74  72]]

 [[ 87 134 146 ... 108 116 157]
  [108 117 144 ... 102  58 122]
  [124 148 106 ...  97 135 146]
  ...
  [ 96 153 111 ... 104 129 154]
  [129 140 100 ...  74 114  97]
  [119 115 160 ... 172  84 148]]

 ...

 [[ 92  96  64 ...  69  83  83]
  [ 85  44  89 ... 115  94  76]
  [ 93 103  91 ...  92  81  75]
  ...
  [ 16 109  81 ...  84  95  20]
  [100  27  89 ...  66 107  48]
  [ 24  67 144 ... 104 115 123]]

 [[ 69  70  74 ...  72  73  75]
  [ 72  72  76 ...  73  75  76]
  [ 74  75  72 ...  72  69  73]
  ...
  [ 72  72  69 ...  72  76  72]
  [ 70  72  73 ...  72  76  67]
  [ 69  72  72 ...  72  71  71]]

 [[ 65 137  26 ... 134  57 174]
  [ 91  76 123 ...  39  63 124]
  [ 81 203 134 ... 192  63 143]
  ...
  [  1 102  96 ...  33  63 169]
  [ 82  32 108 ... 151  75 151]
  [ 12  97 164 ... 101 125  60]]]

y_train:
[0 0 0 ... 0 0 0]

These are my input shapes:

x_train.shape = (5000, 128, 128)
y_train.shape = (5000,)

As you can see, y is just the label and x is just 3D data. Because its a binary classifier, I wanted to build a simple NN with 3 dense layers. This is what I have:

model = Sequential()
model.add(Dense(12, input_dim = 8, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
model.fit(x_train, y_train, epochs=150, batch_size=8)

But due to my input, I am getting this error:

ValueError: Input 0 of layer sequential_4 is incompatible with the layer: expected axis -1 of input shape to have value 8 but received input with shape (8, 128, 128)

How can I fix this problem? Is my NN too simplistic for this type of problem?

tensorflow deep-learning classification machine-learning python
1 Answers

There are multiple issues with your code. I have tried adding separate sections to explain them. Please go through all of them and do try out the code examples I have shown below.

1. Passing the samples/batch channel as the input dimension

You are passing the batch channel as the input shape for the dense layer. That is incorrect. Instead what you need to do is to pass the shape of each sample that the model should expect, in this case (128,128). The model automatically adds a channel in front for the batches to flow through the computation graph as (None, 128, 128), as shown in model.summary() below

2. 2D inputs for Dense layer.

Each of your samples (in this case total of 5000 samples) is a 2D matrix of shape 128,128. A dense layer can not directly consume it without flattening it first. (or using a different layer to be more suited to work with 2D/3D inputs as discussed later).

from tensorflow.keras import Sequential
from tensorflow.keras.layers import Dense, Flatten

model = Sequential()
model.add(Flatten(input_shape=(128,128)))
model.add(Dense(12, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
model.summary()

Model: "sequential_6"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
flatten_3 (Flatten)          (None, 16384)             0         
_________________________________________________________________
dense_5 (Dense)              (None, 12)                196620    
_________________________________________________________________
dense_6 (Dense)              (None, 8)                 104       
_________________________________________________________________
dense_7 (Dense)              (None, 1)                 9         
=================================================================
Total params: 196,733
Trainable params: 196,733
Non-trainable params: 0
_________________________________________________________________

3. Using a different architecture for your problem.

"Is my NN too simplistic for this type of problem?"

It's not about the complexity of the architecture, but more about the type of layers that can process a certain type of data. In this case, you have images with single channels (128,128), which is a 2D input. Usually, color images have R, G, B channels, which end up as (128,128,3) shaped inputs.

The general practice is to use CNN layers for this.

An example of that is shown below -

from tensorflow.keras import Sequential
from tensorflow.keras.layers import Dense, Flatten, Conv2D, MaxPooling2D, Reshape

model = Sequential()
model.add(Reshape((128,128,1), input_shape=(128,128)))
model.add(Conv2D(5, 5, activation='relu'))
model.add(MaxPooling2D((2,2)))
model.add(Conv2D(10, 5, activation='relu'))
model.add(MaxPooling2D((2,2)))
model.add(Conv2D(20, 5, activation='relu'))
model.add(MaxPooling2D((2,2)))
model.add(Conv2D(30, 5, activation='relu'))
model.add(MaxPooling2D((2,2)))
model.add(Flatten())
model.add(Dense(12, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
model.summary()

Model: "sequential_13"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
reshape_5 (Reshape)          (None, 128, 128, 1)       0         
_________________________________________________________________
conv2d_16 (Conv2D)           (None, 124, 124, 5)       130       
_________________________________________________________________
max_pooling2d_16 (MaxPooling (None, 62, 62, 5)         0         
_________________________________________________________________
conv2d_17 (Conv2D)           (None, 58, 58, 10)        1260      
_________________________________________________________________
max_pooling2d_17 (MaxPooling (None, 29, 29, 10)        0         
_________________________________________________________________
conv2d_18 (Conv2D)           (None, 25, 25, 20)        5020      
_________________________________________________________________
max_pooling2d_18 (MaxPooling (None, 12, 12, 20)        0         
_________________________________________________________________
conv2d_19 (Conv2D)           (None, 8, 8, 30)          15030     
_________________________________________________________________
max_pooling2d_19 (MaxPooling (None, 4, 4, 30)          0         
_________________________________________________________________
flatten_9 (Flatten)          (None, 480)               0         
_________________________________________________________________
dense_23 (Dense)             (None, 12)                5772      
_________________________________________________________________
dense_24 (Dense)             (None, 8)                 104       
_________________________________________________________________
dense_25 (Dense)             (None, 1)                 9         
=================================================================
Total params: 27,325
Trainable params: 27,325
Non-trainable params: 0
_________________________________________________________________

To understand what Conv2D layers and MaxPooling layers do, do check out my well-accepted answer on Data Science stack exchange, or check out this blog. But a general way to mentally visualize what it's doing is the following diagram.

enter image description here

user_47559540
Related