Sequential model API in TensorFlow - Model Pipeline Trace

Model Pipeline - Sequential model API

The Sequential model API in TensorFlow helps us build a simple stack of layers for machine learning. We add layers one after another, and the model learns to make predictions from input data.

Data Flow - 5 Stages

1Input Data

1000 rows x 20 columns→Raw numerical features representing samples→1000 rows x 20 columns

[[0.5, 1.2, ..., 0.3], [0.1, 0.4, ..., 0.9], ...]

↓

2Normalization Layer

1000 rows x 20 columns→Scale features to have mean 0 and variance 1→1000 rows x 20 columns

[[-0.1, 0.5, ..., -0.3], [0.0, -0.2, ..., 1.1], ...]

↓

3Dense Layer 1

1000 rows x 20 columns→Fully connected layer with 64 neurons and ReLU activation→1000 rows x 64 columns

[[0.0, 1.2, ..., 0.5], [0.3, 0.7, ..., 0.9], ...]

↓

4Dense Layer 2

1000 rows x 64 columns→Fully connected layer with 32 neurons and ReLU activation→1000 rows x 32 columns

[[0.2, 0.1, ..., 0.4], [0.5, 0.3, ..., 0.7], ...]

↓

5Output Layer

1000 rows x 32 columns→Fully connected layer with 1 neuron and sigmoid activation for binary classification→1000 rows x 1 column

[[0.8], [0.3], [0.6], ...]

Training Trace - Epoch by Epoch

Loss
0.7 |****
0.6 |*** 
0.5 |**  
0.4 |*   
0.3 |*   
     1 2 3 4 5 Epochs

Epoch	Loss ↓	Accuracy ↑	Observation
1	0.65	0.60	Model starts learning, loss is high, accuracy is low
2	0.50	0.75	Loss decreases, accuracy improves
3	0.40	0.82	Model continues to learn well
4	0.35	0.86	Loss decreases steadily, accuracy increases
5	0.30	0.89	Good convergence, model is improving

Prediction Trace - 5 Layers

Layer 1: Input Layer

Layer 2: Normalization Layer

Layer 3: Dense Layer 1 (ReLU)

Layer 4: Dense Layer 2 (ReLU)

Layer 5: Output Layer (Sigmoid)

Model Quiz - 3 Questions

What does the ReLU activation function do in the Dense layers?

A It outputs values between -1 and 1

B It scales all values between 0 and 1

C It changes all negative values to zero and keeps positive values as is

D It sums all inputs without change

Key Insight

The Sequential model API lets us build a simple chain of layers easily. Each layer transforms data step-by-step, and training improves the model by reducing loss and increasing accuracy. Activation functions like ReLU help the model learn complex patterns by introducing non-linearity.