Scribd
Upload
7 views12 pages

Neural Network Basics and Implementation

The document provides an overview of neural networks, including their structure, activation functions, loss functions, and optimizers. It includes code examples for building and training feedforward neural networks, convolutional neural networks, and recurrent neural networks using TensorFlow and Keras. Additionally, it discusses techniques for preventing overfitting, optimizing model performance, and saving/loading models.

Uploaded by

sajid
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views12 pages

Neural Network Basics and Implementation

The document provides an overview of neural networks, including their structure, activation functions, loss functions, and optimizers. It includes code examples for building and training feedforward neural networks, convolutional neural networks, and recurrent neural networks using TensorFlow and Keras. Additionally, it discusses techniques for preventing overfitting, optimizing model performance, and saving/loading models.

Uploaded by

sajid
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Tajamul Khan

@Tajamulkhann
Neural Networks –
Composed of input,
hidden, and output
layers
Activation Functions –
Sigmoid, ReLU, Tanh,
Softmax
Loss Functions – MSE,
Cross-Entropy
Optimizers – SGD, Adam,
RMSprop

@Tajamulkhann
Feedforward Neural Network (FNN)

import tensorflow as tf
from [Link] import
Sequential
from [Link] import Dense

model = Sequential([
Dense(64, activation='relu', input_shape=
(10,)),
Dense(32, activation='relu'),
Dense(1, activation='sigmoid')
])

[Link](optimizer='adam',
loss='binary_crossentropy', metrics=
['accuracy'])
[Link]()

@Tajamulkhann
ReLU (Rectified Linear Unit) –
max(0, x)
Sigmoid – Converts to
probability (0 to 1)
Tanh – Scales between -1 and 1
Softmax – Used in multi-class
classification

import [Link]
as K
def relu(x):
return [Link](0, x)
def sigmoid(x):
return 1 / (1 + [Link](-x))

@Tajamulkhann
Train & Evaluate a Neural Network

history = [Link](X_train,
y_train, epochs=50,
batch_size=32, validation_data=
(X_test, y_test))
loss, accuracy =
[Link](X_test, y_test)
print(f"Test Accuracy: {accuracy
* 100:.2f}%")

@Tajamulkhann
Best for Image Classification &
Object Detection

from [Link] import Conv2D,


MaxPooling2D, Flatten

cnn_model = Sequential([
Conv2D(32, (3,3), activation='relu', input_shape=
(128,128,3)),
MaxPooling2D(2,2),
Conv2D(64, (3,3), activation='relu'),
MaxPooling2D(2,2),
Flatten(),
Dense(128, activation='relu'),
Dense(10, activation='softmax')
])

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

@Tajamulkhann
Best for Sequential Data like
Time Series & NLP

from [Link] import


SimpleRNN, LSTM

rnn_model = Sequential([
LSTM(50, return_sequences=True,
input_shape=(100, 1)),
LSTM(50),
Dense(1)
])

rnn_model.compile(optimizer='adam',
loss='mse')

@Tajamulkhann
Best for NLP & Vision Tasks

from transformers import


TFAutoModel

bert_model =
TFAutoModel.from_pretraine
d("bert-base-uncased")

@Tajamulkhann
Prevent Overfitting – Dropout,
L1/L2 Regularization

from
[Link]
import Dropout

[Link](Dropout(0.3))
# Dropout 30%

@Tajamulkhann
Optimize Model Performance

from keras_tuner import


RandomSearch

tuner = RandomSearch(model,
objective='val_accuracy',
max_trials=5)
[Link](X_train, y_train,
epochs=10, validation_data=
(X_test, y_test))

@Tajamulkhann
Save & Load a Model

[Link]("my_model.h5"
)

loaded_model =
[Link].load_mod
el("my_model.h5")

@Tajamulkhann
Follow for more!

You might also like