WORKSHOP INLINE WITH ICW-HDDA-X 2020
These workshop activities are related to ICW-HDDA-X. Workshop topics are selected according to the theme and seminar track, also it is connected with the research group in the Mathematics department of the host institution.
Each ICW-HDDA-X participant can attend the workshop by filling in the registration form and selecting the workshop topic as needed.
Workshop registration : https://bit.ly/WorkshopRegistrationICWHDDA10
- Toward Data Science: Application of Optimal Transport Theory for Unsupervised Generative Adversarial Networks
Toward Data Science:
Data Science is multi-disciplinary field that uses many scientific methods, processes, techniques and algorithms to learn patterns and extract insights from many structural and unstructured data. Data science is related to data mining, machine learning and deep learning. It is a unified concept of statistics, data analysis, pattern recognition, machine learning and domain knowledge. It uses many techniques and theories from mathematics, statistics and information theory. In this session, we will explain what is data science and some of our programs in data science center of University of Indonesia.
Optimal transport (OT) theory is gradually getting more attractions from AI researchers as a powerful and essential method to compare probability measures, which is very important aspect in machine learning techniques. OT is applicable not only for discriminative models, but also in the emerging unsupervised generative models like Generative Adversarial Network (GAN). In this session, we will explain and demonstrate how optimal transport based divergence can be used in common GAN models to generate realistic data samples.
- Speakers: Risman Adnan (Data Science Center(DSC) Universitas Indonesia & Samsung Research & Development Indonesia(SRIN))
- Coordinator: Alhadi Bustamam (Department of Mathematics, and Data Science Center (DSC), Universitas Indonesia)
- Data Wrangling and Scraping
Data is now everywhere at any time. However, not all data is readily available for analysis. Cleaning and pre-processing data are important steps in data analytics but often neglected. In this workshop, you will learn and practice how to turn unrefined, unstructured and messy data into something useful for analysis using R.
- Mochamad Kautzar Ichramsyah (Senior Data Analyst, Traveloka)
- Danardono (Department of Mathematics, Universitas Gadjah Mada)
- Coordinator: Danardono (Department of Mathematics, Universitas Gadjah Mada)
- Gaussian Copula processes on time series data
In this workshop we will learn how to model time series data using function spaces. The time series data can be seen as a set of function values which are generated by an underlying function. This underlying function which is unknown will be assumed to be an element of a special function space. In this workshop we will consider the function spaces which are derived from Gaussian Copula processes. To apply this approach, some financial time series data will be considered. The results will be compared with other approaches to show the performance of Gaussian Copula processes.
- Speakers: Sapto Wahyu Indratno (Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung
- Coordinator: Novry Erwina (Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung)
- Islamic Mathematical Finance
The development of Islamic investment in many countries, including Indonesia, is still in slow growth. There are still reluctance for implementing a profit-sharing investment scheme due to its complex procedure on determination of the optimal portion or nisbah of profit sharing. In this workshop, we determine theoretically the optimal nisbah on Musyarakah contract between traditional market traders and an investor. In this contract, the capital of the business is contributed both by investor and traders, so the profit and loss resulted from the business activities will be shared between two parties. Firstly, we will learn basics of Islamic Finance. In the modelling process, we will observe the data of trader’s profit, generating a simulation data, determination of an objective function, and running an optimization method using EXCEL and MATLAB software.
- Speakers: Novriana Sumarti (Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung)
- Coordinator: Rudy Kusdiantara (Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung)
- Spatio Temporal Modeling for Media Social Data
In this workshop the participants will take the opportunity to study more detail about Spatio Temporal Modeling for Media Social Data, especially to discuss how to access the trending topic on media social data such as Twitter. For the case study, we use the Covid-19 data using Time Series, Spatial, and Spatio Temporal Models for description and prediction of real phenomena using R Software. The workshop also part of the activity of the International Consortium Research Innovation and Staff Exchange on Social Media Analytics (RISE_SMA) for schema Horizon 2020 funded by the European Union for the year 2019-2022 with University Duisburg Essen as a Coordinator.
- Atje Setiawan Abdullah (Department of Computer Science, Universitas Padjadjaran)
- Budi Nurani Ruchjana, Annisa Nur Falah, Mutik Alawiyah, and Devi Munandar (Department of Mathematics, Universitas Padjadjaran)
- Eddy Hermawan (National Institute of Aeronautics and Space of Indonesia)
- Coordinator: Budi Nurani Ruchjana (Department of Mathematics, Science and Technology Studies Center, Faculty of Mathematics and Natural Sciences, Universitas Padjadjaran)
- Math Learning Media with VBA for Excel
- Martin Bernard, Wahyu Hidayat (IKIP Siliwangi)
- Rully Charitas Indra Prahmana (Universitas Ahmad Dahlan)
- Koordinator: Wahyu Hidayat (IKIP Siliwangi)