Young Innovate

Divided into 3 categories:

Students have to develop algorithm that can predict target with highest accuracy in financial, such as :

• Stock market prediction
• Housing stock prediction
• FOREX stock prediction
• example : Tenaga Nasional Berhad profit and loss prediction

Students have to develop algorithm that can analyze the target using camera to solve problems in agriculture, such as:

• Plant health monitoring and prediction
• Fruit maturing process monitoring and prediction
• Plant enemy detection and decision making
• example : Detecting oil pam desease by using image

Students have to develop algorithm that can analyze the target using camera to solve problems in manufacturing, such as:

• Product defect detection
• Manufacturing plant efficientcy monitoring and prediction
• Robot arm movement optimization and decision making
• example : Real-time detection and optimization of robot arm

Process Flow

Final Step :

Decision Making

Easy Aid Python-API for Machine Learning with Cloud

API

Application Programming Interface

Cloud

Cloud Servers

Easy Aid Python-API for Machine Learning with Cloud

What can API-Cloud can do?

• Global common programming language Python
• Mathemathics, Science, and Engineering eco-system
• Scientific computing package
• Outstanding data visualization
• High Perfomance
• Rich statistical models, data exploration tools
• Data mining and data analysis
• Machine Learning algorithms
• Highly automated cloud
• Easy deployment
• Data Scientist Training

Iterative Development Process

Multiple Data

Model Development

API Programming with Cloud

Model Tuning

Evaluate Model

Easy Aid Python-API for Machine Learning with Cloud

List of algorithm in API....

Prediction / Regression

Classification

Optimization

• First Order Optimization Algorithms
• Stochastic Gradient Descent
• Adagrad
• ADAM
• AdaBoost
• AdaDelta
• Decision Tree

• Deep Learning
• Bayesian Linear Regression
• Boosted Decision Tree
• Random Forest Quantile Regression
• Long-short-term-memory (LSTM ) time-series
• Recurrent Neural Network (RNN)
• Multilayer Perceptron (MLP)

• Convolutional Neural Network (CNN)
• Support Vector Machine (SVM)
• Mutiple-class decision forest
• Bayesian Point classicification
• K-Neighbours Classifier
• Stochastic Gradient Descent Classifier
• Discriminatn Analysis

Module 1 : Smart Economic

1

2

3

4

• For this category, we are focusing straight to

"Time-Series Algorithm" , which are divided into 4

months project-based.

Timeline for 4 months

• Investment Prediction

• Energy profit and loss

prediction

• Housing Stock Prediction

• Money-exchange

Stock Prediction

#the level of difficulty are increasing

Time-Series Forecasting

• Time-series Nomenclature

• Components of Time-series

• Supervised Machine Learning

• Sliding Window in Unvariate Time Steps

• Sliding Window in Multiple Steps

1st Month -

Data Preparation (Pre-Processing)

• Load and Explore Time-series Data

• Basic Feature Engineering

• Time-series Visualization

• Resampling and Interpolation

• Power Transforms

• Moving Average Smoothing

2nd Month - Evaluate Models

• Backtest Forecast Models

• Forecasting Perfomance Measures

• Persistence Model for Forecasting

• Visualize Residual Forecast Errors

• Reframe Time Series Forecasting Problems

3rd Month - Forecast Models (Projects)

Example Project: Monthly Sales of Housing Stock

a) Overview

b) Problem Description

c) Test Harness

d) Persistence

e) Data Analysis

f) ARIMA Models

g) Model Validation

4th Month - LSTM for Stock Prediction

• What are LSTM?

• Prepare data for LSTM

• Develop LSTM in Keras

• Models for sequence prediction ( Vanilla, Stacked, CNN, LSTM, Encoder-Decoder)

• Finalize a LSTM model, Save LSTM models, Make predictions on New Data

Module 2 : Smart Farming

1

2

3

4

• For this category, we are focusing straight to

"Deep Learning algorithm" , which are divided into 4

months project-based.

Timeline for 4 months

• Fruit maturing classification

• Camera integration

with Raspberry Pi

• Plant type

classification

• Plant enemy detection

#the level of difficulty are increasing

Module 3 : Smart Manufacturing

1

2

3

4

• For this category, we are focusing straight to

"Deep Learning algorithm" , which are divided into 4

months project-based.

Timeline for 4 months

• Robotic arm optimization

• Product Classification

• Robotic Arm

Integration &

Programming

• Robotic arm optimization with decision making

#the level of difficulty are increasing

Raspberry Pi Computer Kit:

- Raspberry Pi 3

- Memory Card

- VGA / HDMI cable

- Power Supply

- Casing

- Keyboard and mouse

- with additional electronic kit

Raspberry Pi Computer Kit:

- Raspberry Pi 3

- Memory Card

- VGA / HDMI cable

- Power Supply

- Casing

- Keyboard and mouse

- Camera module and additional sensors

- with additional electronic kit

Raspberry Pi Computer Kit:

- Robotic Arm

- Raspberry Pi 3

- Memory Card

- VGA / HDMI cable

- Power Supply

- Casing

- Keyboard and mouse

- Camera module and additional sensors

- with additional electronic kit

UniKL MSI EV Interest Group

Prof. Dr. Abu Talib

Head of Campus / Dean

Dr. Muhamad Husaini

bin Abu Bakar

Research & Innovation Section

Smart Manufacturing

Machine Learning

Smart Farming

Smart Economic

Prof. Dr. Sazali bin

Yaacob

Electrical, Electronics

& Automation Section

Mohd Zaki bin

Abdul Razak

Smart

Manufacturing

Section

Prof. Dr. Azmi

Electrical, Electronics

& Automation Section

Saharul bin Arof

Electrical, Electronics

& Automation Section

Nuraida Md. Hassan

Electrical, Electronics

& Automation Section

Tajul Adli bin

Abdul Razak

Mechanical

Section

Siti Lydia

binti Rahim

Manufacturing

Section

Mohamed Yusof

bin Radzak

Head of Section

Workshop Section

Shahruzaman bin

Sulaiman

Smart

Manufacturing

Section

Amir Shauqee bin

Abdul Rahman

Electrical, Electronics

& Automation Section

Dr. Rahim Jamian

Manufacturing

Section

Mohamad Sabri

bin Mohamad Sidik

Mechanical

Section

Mohd Nurhidayat

bin Zahelem

Mechanical

Section