Scribd
Upload
9 views2 pages

Week 2 Project

The document outlines two case studies on modeling neural networks: one for house price estimation and another for employee performance evaluation, each utilizing a single-layer neural network. It also describes two problems involving multi-layer neural networks for predicting energy consumption and academic performance, detailing the input parameters, network architecture, and weights for each model. ReLU activation function is applied in the hidden layers of the multi-layer networks.

Uploaded by

pramodk8008485289
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
9 views2 pages

Week 2 Project

The document outlines two case studies on modeling neural networks: one for house price estimation and another for employee performance evaluation, each utilizing a single-layer neural network. It also describes two problems involving multi-layer neural networks for predicting energy consumption and academic performance, detailing the input parameters, network architecture, and weights for each model. ReLU activation function is applied in the hidden layers of the multi-layer networks.

Uploaded by

pramodk8008485289
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

Micro Project-2

Session-3: Modeling a Single-Layer Neural Network Using Matrix Multiplication


Case Study 1: House Price Estimation
Predict the house price score using the following factors:
• 𝑥1 : House size (in 100 [Link])
• 𝑥2 : Distance from city center
• 𝑥3 : Age of the house
Hidden Linear Layer (2 Neurons): The input vector, weight matrix, and bias vector are given
by:
7
0.6 −0.2 −0.1 0.5
𝑋 = [3] , 𝑊 = [ ],𝑏 = [ ]
0.4 −0.3 −0.2 0.2
2

and Final Output Layer (1 Neuron): Weight and Bias 𝑊2 = [0.6 0.4], 𝑏2 = 0.5.

Case Study 2: Employee Performance Evaluation


Predict an employee performance score based on:
• 𝑥1 : Years of experience
• 𝑥2 : Skill assessment score
• 𝑥3 : Training hours per month
Hidden Linear Layer (2 Neurons): The employee data and network parameters are:
5
0.3 0.4 0.2 0.8
𝑋 = [7 ] , 𝑊 = [ ],𝑏 = [ ]
0.2 0.5 0.1 0.4
4
Final Output Layer (1 Neuron): Weight and Bias 𝑊2 = [0.7 0.3], 𝑏2 = 0.4.

Session-4: Modeling a Multi-Layer Neural Network Using Weight and Activation


Matrices
Problem 1: Modeling Energy Consumption Using a Multi-Layer Neural Network
An energy management system aims to predict the daily energy consumption index of an
industrial unit based on the following parameters:
• 𝑥1 : Operating hours per day
• 𝑥2 : Machine load factor
• 𝑥3 : Ambient temperature
The system is modeled using a multi-layer neural network with:
• One input layer (3 neurons),
• Two hidden layers (each with 2 neurons),
• One output layer (1 neuron).
The network parameters are given as follows:
6
0.5 0.3 0.2 0.4
Input Vector 𝑋 = [4], Hidden Layer 1: 𝑊1 = [ ] , 𝑏 = [ ],
0.2 0.6 0.1 1 0.3
5
0.6 0.2 0.2
Hidden Layer 2: 𝑊2 = [ ],𝑏 = [ ]
0.3 0.7 2 0.4
Output Layer: 𝑊3 = [0.7 0.3], 𝑏3 = 0.5. Assume the ReLU activation function is applied
after each hidden layer.
Problem 2: Multi-Layer Neural Network for Academic Performance Prediction
Problem Statement
A university uses a neural network model to estimate a student success score based on:
• 𝑥1 : Weekly study hours
• 𝑥2 : Attendance score
• 𝑥3 : Assignment performance score
The neural network architecture consists of:
• Input layer with 3 neurons,
• Two hidden layers with 2 neurons each,
• Output layer with 1 neuron.
The network parameters are defined as:
7
0.4 0.5 0.1 0.6
Input Vector: 𝑋 = [6], Hidden Layer 1: 𝑊1 = [ ] , 𝑏1 = [ ].
0.3 0.4 0.2 0.4
8
0.5 0.3 0.3
Hidden Layer 2: 𝑊2 = [ ] , 𝑏2 = [ ].
0.4 0.6 0.2
Output Layer 𝑊3 = [0.6 0.4], 𝑏3 = 0.5. Assume ReLU activation is used in all hidden
layers.

You might also like