Data Science Academy Camp 1: Beginning of the journey
COMPFEST 13, Jakarta — Camp 1 of Data Science Academy has been held from August 16, 2021, to August 21, 2021. All of the sessions in this Camp were held online via Zoom Meeting. Participants from the Data Science Academy are the ten best teams previously chosen by COMPFEST. Camp 1 of Data Science Academy primarily discusses the early steps of data science. Let’s see the excitement!
Day 1 — Introduction to Data Science Workflow Business
The Camp started with the opening remarks from Raja Aldwyn Sihombing, Project Officer of COMPFEST 13, and Radhiansya Zain Antriksa Putra, Vice Manager of Academy COMPFEST 13. From there on, we dove straight into our first speaker session with Ezra Aminanto, Data Scientist from Jakarta Smart City. On this occasion, Ezra explained the Introduction to Data Science Workflow Business.
This session started with a discussion about the definition of data science. Ezra stated that data science is an interdisciplinary field that builds on statistics, informatics, computing, communication, management, and sociology to study data and its environment, in order to get insights and output that can be used for decision making. Ezra also explained the three ways to think about data — data-driven, data-informed, and data-aware — and its real case in the industry.
Furthermore, Ezra talked about the data science methodology. The process begins with understanding the business purpose and doing an analytic approach, followed by the preparation steps; data requirements, data collection, data understanding, and data preparation. Then the process continues to the modeling process, followed by the evaluation, deployment, and feedback process. “This process is not always linear, it can jump from one stage to another if needed,” Ezra stated.
Moreover, Ezra went on to discuss two ways to think about doing data science; Top-Down and Bottom-Up. To end the presentation, Ezra explained the skills needed in each step of the data science methodology.
Day 2 — Pre-Processing & Cleansing
The second day of Data Science Academy kicked off with a presentation from Evan Budianto, a data scientist from Data Science Indonesia. The topic he presented to the audience was “Pre-Processing & Cleansing”.
To start his presentation, Evan greeted the participants warmly. He then went on to discuss the definition of pre-processing and cleansing and why this process is necessary. Pre-processing and cleansing is a preparation process that involves transforming raw data into a form that is more appropriate for modeling, to improve the performance of machine learning algorithms. “Better data beats more data, even though more data beats less data,” Evan stated. In addition, Evan also explained the steps in doing Preprocessing & Cleansing; (1)Handling missing values, (2)Feature scaling, (3)Feature Engineering, (4)Feature encoding, and (5)Dimensionality Reduction.
The Camp then moved on to the hands-on session. In this session, Evan demonstrated the process of pre-processing and cleansing in a real study case. Evan showed comprehensive steps transforming raw data into feasible data for the data analysis process.
Day 3 — Exploratory Data Analysis
On the third day, Irfan Dwiki Bhaswara, a data scientist from Jakarta Smart City, presented a topic about Exploratory Data Analysis (EDA).
Irfan stated that EDA is a step in performing initial investigations on data by discovering patterns, spotting anomalies, and testing hypotheses, to identify important features, check data quality, and test assumptions. And in this way, data-driven insights can be delivered to business stakeholders.
Furthermore, Irfan explained the role of graphics in EDA. That, he said, graphical tools are not just tools that can be used in EDA. However, these procedures are the fastest path to getting insights in terms of testing hypotheses.
Next was the hands-on session. In this session, Irfan demonstrated the EDA process using tools from python package and Google Data studio.
Day 4 — Exploratory Data Analysis 2.0 & Case Study
The last day of Camp 1 brought Hansen Wiguna, a Business Analyst at Jakarta Smart City, to be the speaker and discuss the topic of Exploratory Data Analysis 2.0.
To start his presentation, Hansen explained the definition of EDA. Hansen stated that data scientists have got to thoroughly understand datasets, in order to gain the best solution to the problem. Hansen then explained analysis questions, which are some common questions to help data scientists thoroughly understand datasets. From that questions, data scientists should identify the relevant questions to solve the problem. EDA techniques can be used to support this process.
Afterward, he explained the techniques of EDA; graphical techniques, and quantitative techniques. Graphical techniques can be performed with; univariate, multivariate, time series, 1 factor, multi-factor, and regression, whereas quantitative techniques can be performed with interval estimates and hypothesis tests. Hansen then went on to discuss 6 plots, which is a tool to validate data.
Hansen also shared his experience as a business analyst at Jakarta Smart City. He showed the processing of data and data products in a government institution. “The world needs people who can cleverly solve complex problems, especially in a government institution because the problems here involve and affect a large number of people“, Hansen stated. The speaker session ended with a quick QnA. The Camp then moved on to the hands-on session.
The last day of Camp 1 resumed after a break, the Camp went on to the case study session. The participants were divided into 10 teams to solve the case study in the breakout rooms for 150 minutes. After the teams finished the case, the session continued with a presentation of the discussion result.
Interview Session!
After the closing of Data Science Academy camp 1, we had the opportunity to interview Hansen Wiguna, a Business Analyst at Jakarta Smart City. To start, Hansen shared his interest in Jakarta Smart City. According to him, working in data analytics at a government institution is very interesting because the problems that are solved are problems that involve and affect a large number of people. From there on, he had the opportunity to collaborate with many parties and represent the provincial government at various events. Hansen also shared tips for learning data science to become a competent data scientist. “Know your potential in the 3 domains of data science (business, statistics, and programming) and continue to hone those skills. Don’t be afraid to learn because nowadays data scientists come from various backgrounds, and now there are many learning resources, such as this Data Science Academy,” said Hansen.
Not only that, we also had the opportunity to interview Bryan, Zulfah, and Timot from the HAKATON! team. They shared their memorable experiences while attending the Data Science Academy. HAKATON! said that the case study is one of it because they can meet and talk with other team participants from various backgrounds while honing their skills by exploring the datasets that have been provided. By joining the Data Science Academy, they hope that they will gain a lot of new knowledge from the final project and mentoring session, as well as getting new friends who have the same interest in data science.
See you in the next series of events at COMPFEST! Stay tuned for information about COMPFEST through our social media accounts on Twitter @COMPFEST, Instagram @COMPFEST, Facebook COMPFEST, LinkedIn COMPFEST, and website compfest.id (Editorial Marketing/Visi).
