Applied Data Science
Duration: 2-days
Title:
Applied Data Science
Summary:
This tutorial will attempt to introduce the participant to a holistic approach to practical Data Science activities: Data Wrangling, Model Development, Model Evaluation and Operationalization of Predictive Models. The tutorial will involve both slide presentations as well as hands-on practical work on the activities associated to the Life-Cycle in Predictive Models. Exercise (practical hands-on) work on data preparation, modelling, evaluation and operationalization of the machine learning algorithms in Section B and Section C will be done using the Python Jupiter Notebook and BigML platform. (Basic Account to this platform will be sufficient. There is no cost associated to creating a basic account). Sessions on each of the ML algorithms will be accompanied with a Use Case that will engage the participants in practical hands-on sessions. The tutorial will cover the following topics:
(A) INTRODUCTION
a. Artificial Intelligence in Data Science
b. Data Wrangling
c. Predictive Model Building
d. Model Evaluation
e. Operationalization
f. Monitoring & Model Revision
(B) SUPERVISED MODELING
a. Decision Trees
b. Ensembles
c. Linear Regression
d. Logistic Regression
e. DeepNet
f. Time Series
g. OptiML (Auto Hyperparameter Optimization)
h. Fusion (Model Fusion)
(C) UNSUPERVISED MODELING
a. Clustering
b. Anomaly Detection
c. Association Discovery
d. Topic Modelling
e. Principal Component Analysis
(D) CASE BASED REASONING
(E) SEMANTIC KNOWLEDGE GRAPH
(F) GRAPH ANALYTICS
Requirements (optional):
(1) Internet Connection
(2) Laptop
Who should attend:
(1) Young Data Scientist
(2) People interested in Applied Data Science
Bio-Data:
Dr. Dickson Lukose
|
 |
With over 25 years of experience in professional consulting services, applied research and development on Intelligence Systems, Dr. Lukose now leads a team of Data Scientist doing Big Data Analytics (BDA). Prior to joining GCS Agile, Dr. Lukose was the Senior Director of the Artificial Intelligence (AI) Lab in MIMOS Berhad (Malaysia) developing AI software for Big Data Analytics using various AI technologies, specifically Machine Learning, Deep Learning, Semantic Technology, Semantic Data Lake, large scale Social Network Analytics and Natural Language Processing. Over his tenure at MIMOS, Dr Lukose has completed numerous national level projects for both the Private and Public sectors. Dr. Lukose joined MIMOS Berhad in June 2007 as Chief Research Director and Director of Artificial Intelligence Centre. Earlier in his career, Dr. Lukose worked as Principal Knowledge Engineer with Mindbox Inc. (USA), responsible for architecture and development of high performance Artificial Intelligence software for the Financial Sector. He holds a BSc (Hons) and a PhD from Deakin University (Australia).
Contact:
dickson.lukose@gmail
