Projects
This is several of my data project I created during my path to acquire data knowledge. In this project I showed my interest in how generate story using data and to make better decision with it.
4 May 2020
-
The objective is to determine rice field proportion and location that covers particular region in Indonesia, which in this project is Banyuwangi
-
Use object detection algorithm (Single Shot Multibox Detector) to achieve those objectives.
-
Use ArcGIS pro to label each object into four classes that constitutes satellite raster imagery
-
Use ArcGIS pro to augment and pre-process the image data to make it compatible for model fitting and training.
-
Use python module named arcgis.learn which is included in ArcGIS pro built in python module to build SSD model algorithm.
_edited_.jpg)
Covid-19 Analysis
9 March 2020
-
This project is made when Covid-19 pandemic was still in early phase, thus the data is still limited and cannot be used as real guidelines.
-
The objective is just to do simple data exploration and visualization using Covid-19 case data that had been published by John Hopkins University.
-
Data visualization is generated to know the number of Case Fatality Rate and Recovery Rate through using GGPLOT, the well known R-package for creating compelling data visualization.
-
Prediction modeling is done by using FB Prophet machine learning algorithm to know the number of cases and death in the extend of next month, assuming that the growth rate has exponential nature.
Vaccine Receiver Predictor - On Going
27 May 2020
-
The project is part of the challenge held by DrivenData called "Flu Shot Learning". The purpose of this challenge is to predict the probability of receiving vaccine from each of survey respondents based on the answer they gave to the medical authority.
-
There are some statistic works and data cleaning process before feeding it into machine learning models, such as examining the collinearity of each variable and imputing the missing data using several method, such as logistic regression and median.
-
The preliminary model that is presented is using logistic regression, since the purpose is just to determine the most important variable toward the model.
-
There are several works to done such as creating the model to predict the probability of receiving vaccine and several other statistical examination in order to create the model as robust as possible.
Shopee Code League 2020 - On Going
16 June 2020
-
This project contains my take on several code league challenges that is held by Shopee and is still running until now.
-
The challenges span on wide ranges of topics including order brushing detection, product detection, title translation, logistics, sentiment analysis, marketing analytics, and algorithm problem.
