Scribd
Upload
102 views3 pages

Introduction To Fully Connected NN - Assignment

The document discusses developing a fully connected neural network to classify images from the Fashion MNIST dataset into 10 fashion categories. It covers data preprocessing, designing and training a model, and evaluating performance on training and validation data.

Uploaded by

sayantani 11
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
102 views3 pages

Introduction To Fully Connected NN - Assignment

The document discusses developing a fully connected neural network to classify images from the Fashion MNIST dataset into 10 fashion categories. It covers data preprocessing, designing and training a model, and evaluating performance on training and validation data.

Uploaded by

sayantani 11
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
You are on page 1/ 3

t

aa
Module 4: Introduction to Fully
Connected NN - Assignment
lliP
te
In

Contact us: [email protected] / © Copyright Intellipaat / All rights reserved


Artificial Intelligence Certification Course

Problem Statement:

Fashion MNIST is a dataset comprising 60,000 training images and 10,000


testing images, each depicting various fashion items in a 28x28 grayscale format.
The dataset encompasses ten distinct categories of apparel. The primary aim of
this assignment is to develop a fully connected neural network (FCNN) that can
accurately classify these images into the correct fashion categories.

Objectives:

t
Data Visualization and Preprocessing:

aa
● Visualize the distribution of different fashion items in the dataset to
understand the data better.

● Normalize the image pixel values for effective model training.


lliP
● Convert the categorical labels into a one-hot encoded format to suit the
neural network’s output layer.

Model Architecture and Development:

● Design a fully connected neural network with appropriate input, hidden,


and output layers.
te

● Ensure the model is complex enough to capture the underlying patterns in


the data without overfitting.

Model Training:
In

● Train the model on the preprocessed training data, using categorical


cross-entropy as the loss function, stochastic gradient descent as the
optimizer, and accuracy as the evaluation metric.

● Utilize a validation split to monitor the model’s performance on unseen


data during training.

Contact us: [email protected] / © Copyright Intellipaat / All rights reserved


Artificial Intelligence Certification Course

Performance Evaluation:

● Assess the model’s performance using both training and validation


datasets.

● Visualize the training and validation loss and accuracy to identify any signs
of overfitting or underfitting.

t
aa
lliP
te
In

Contact us: [email protected] / © Copyright Intellipaat / All rights reserved

You might also like