Hello I'm

Imesh Ekanayake

Computer Engineer / Data Scientist


About Me

A committed, adaptive and Solution orientated leader who is passionate in research and innovations while keeping continues learning and imagination as the best asset. I believe DATA can speak about the present and future more accurately than any expert.

  • Data Mining
  • Machine Learning
  • NLP
  • Public Speaking
  • Compute Vision
  • Algoritham Developer
  • Business Development
Downlaod CV

What I do

img04

Machine learning Applied research

Apply machine learning algorithms on various datasets, and combine differnt algorithms and build ensemble systems based on the advantages of each model.

Classical ML, Time Series, CNN & RNN
img04

Neural Networks and architecture design

Design and modify neural network architectures to obtimes the learning based on the scarce resources.

Dense Net, UNet, Resnet

img04

Algorithm Development

Apply and modify data structures and develop algorithms to complete the task optimally. Good at Dynamic Programming, Recursion algorithms, Greedy algorithms and sorting.

Sorting, Search, Graphs, DP, Recursion, Greedy
img04

Business Development

Business idea development, startup building and negotiation are practicing while making stratgic plans to obtain the trust and increase the qulity of the output.


Start up, Entrepreneur

Technical Skills

Python
R
C
Java
Microsoft Azure
Google Cloud Platform
SQL
Spark

Professional Skills

  • Critical Thinking
  • Problem Solving
  • Team-wrok & Management
  • Communication & Presentation
  • Leadership
  • Time Management

Analytical Skills

Classical Machine Learning
Neural Network and Deeplearning
Algoritham Development
Computer Vision
Natural Language Processing
Feature Engineering

Education

Computer EngineeringUniversity of Peradeniya

2016-2021

Bachelor of the Science of Engineering (BScEng)
Specialization: Computer Engineering

CIMA Operational Level

2016-2021

Got a clear understanding of the key aspects of businesses and to align ideas with businesses to make them commercial

Proffesional Experience

Casual Instructor University of Peradeniya

2020-feb (Present)

Department of Computer Engineering
University of Peradeniya

Research Assistant University of Moratuwa

2015

Research on an animal detecting system for Hambanthota Open Park
Research Assistant (Under Dr.Chandana Gamage) Department of Computer Science Engineering

Voluntering AIESEC

2017-2019
Responsibility :
  • Volenteered as an Instructor for Computer Idea development and Idea presenting in Bologna, Italy
  • Worked in the Business Development piller from the member role until the vice-precident
  • Partnership Development Specilist in the biggest AIESEC international conference (70M LKR Conference)

Proffesional Certifications

TensorFlow Specialization deeplearning.ai

Intro to TensorFlow :- Issued May 2020 Credentials
NLP :- Issued May 2020 Credentials
CNN :- Issued May 2020 Credentials
Time Series :- Issued May 2020 Credentials

Google Data Engineer Google cloud platform

2020
Data Pipelining :- Issued April 2020 Credentials
Data Storing and Distributed Processing :- Issued April 2020 Credentials
Data Management :- Issued May 2020 Credentials
Building Resilient Streaming :- Issued April 2020 Credentials
Smart Analytics :- Issued May 2020 Credentials
Image understanding :- Issued Jun 2020 Credentials

AI for Medicine deeplearning.ai

2020
Diagnosis :- Issued May 2020 Credentials
Prognosis :- Issued May 2020 Credentials
Treatment :- Issued May 2020 Credentials

Machine Learning Stanford Online

2019
Machine Learning :- Issued Sep 2019 Credentials

Recent Portfolio

  • All Categories
  • Machine Learning
  • Embeded
  • Software

Credit Approval Prediction

Jan 2020 – May 2020

Initially the raw data has been taken and passed through the prepossessing which consist of encoding the categorical data and replacing the missing values. Thereaǒter, data has been visualized in many different ways using dimensionality reduction methods, removed out liers and checked the correlation matrix and checked the features with the domain understanding. Next, 11 different models were trained using Classical Machine Learning models, A neural qnetwork has been trained with the full data set and the PCAed data set and finally two models (XGB classifier and CatB classifier) have trained using genetic Optimization for 100 generations. Finally based on the F1 score and accuracy the models were selected which are XGB with granitic optimization and the neural network trained with full data set.

  • Pandas
  • Numpy
  • sk-learn
  • Keras
  • XGBoost
  • genetic Optimization

Distribution of the original dataset

Dimensionality reduction from TSNE .

Chronic Kidney Disease Prediction

Dec 2019 – Feb 2020

Chronic Kidney Disease (CKD) or chronic renal disease has become a major issue with a steady growth rate. A person can only survive without kidneys for an average time of 18 days, which makes a huge demand for a kidney transplant and Dialysis. It is important to have effective methods for early prediction of CKD. Machine learning methods have been shown to be effective in CKD prediction. This work proposes a workflow to predict CKD status based on clinical data, incorporating data prepossessing, a missing value handling method with collaborative filtering and attributes selection. Out of the 11 machine learning methods considered, the Extra tree classifier and Random forest classifier are shown to result in highest accuracy and minimal bias to the attributes. The research also considers the practical aspects in data collection and highlights the importance of incorporating domain knowledge when using machine learning for CKD status prediction.

  • Classical ML Models
  • Feature Engineering
  • Imputing
  • Feature impotance analysis
  • Model Selection

The Distribution of the original variables are above.

Bicycle and scooter sharing system

May 2019 – Nov 2019

As most of the universities have a wide area of land, transportation within the university causes time to waste, accidents, congestion because of using the private vehicles, parking problems and the energy consumption related to the mobility of workers and students of the universities. The bicycle-sharing programs have received increasing attention in recent years with initiatives to increase bike usage, better meet the demand of a more mobile public and lessen the environmental impacts of our transportation activities. So the project aims to introduce an automated bike-sharing system to minimize the above impacts while evaluating the mobility patterns of academic campuses and assessing the energy consumption and pollutant emissions produced by the universities. This system provides the users to unlock the chosen bicycle in the substations via a mobile app and start riding, check the availability of bicycles and authorized people to track the path of rides of all users. This is a system which provides a facility to borrow bicycles or scooters which parked in a dock and the system consists of a Web application for Admin, Mobil application for users or messaging system to get a code and unlock the bike.

  • Android
  • Arduino
  • GSM
  • GPRS
  • Web
  • Google API
  • Solidworks and CAD



The above video is the Demo presentaion of the project.

Above video descibes the locking mechanism of the dock.

Finite Element Analysis Software

Jan 2019 – Jun 2019

2D and 3D structure linear analysis and 2D structure non-linear analysis implementation to analyze structural stability of the buildings and bridges. The
Theoretical Research : Sameera Hippola
Project Supervisor : Dr. K.K.Wijesundara.

  • Python
  • Numpy
  • Structural Engineering Basics

A comparison on currently using software and the model designed.

Stock market Platform

Nov 2018 – Dec 2018

descriptionThis project is about creating a platform to bidding on stocks and by using the data predict the ups & downs of the price as an additional feature.

  • Time Series
  • Arima Model
  • LSTM
  • Python

Above you can find a predicted graph.

Sinhala Letter recognition

Aug 2018 – Nov 2018

This is aboit letter recognition from Machine learning. Sinhala has 60 complex main letters and 20+ special characters. which makes every letter different from each other.

  • CNN
  • Tensorflow
  • Python

Sinhala letter segmentation.

8 bit Computer

Mar 2018 – Jul 2018

Embedded project to created an 8-bit computer using Lowe level component (Logic Gates , Multiplexes , EROM and etc.)

  • Gate level designing
  • Mux and Demux
  • Encorders
  • Counters

The project was designed from Verilog .

Data Over Sound

May 2018

low energy short rang networking vie ultra sound waves . which can use to communicate with out using any radio waves (Bluetooth, Wi-Fi or cellular data .)

  • Android
  • Sound waves
  • Noice Canceling
  • Encryption
  • PC software

The above image shows the process of transfering data.

Flow Rate meter for Dengue patients

May 2017 – Nov 2017

Because of the lack of automated urine measuring instruments in the government hospitals, created an low cost flow rate meter to measure the amount of urine passed in a Dengue patients. which used thermal inductance of the liquid and the used machine learning to rectify the errors , used a reputation model to make the things work.

  • Arduino
  • Electronic Sensors
  • Machine Learing

All variations are organized separately so you can use / customize the template very easily.

Blog Posts

Breast Cancer Analysis

It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout

Read More

Credit-card Approval and Categorization
(UCI - Dataset)

It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout

Read More

Credit Approval prediction with Gentic Optimization

It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout

Read More

Imputing Missing Values

It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout

Read More
Colors :