AsiaQuest Indonesia
AsiaQuest Indonesia AsiaQuest Indonesia

News / Blog

The No.1 WEB development in Indonesia. AsiaQuest Indonesia will share news, events, and blog of WEB utilization in Indonesia. We post in English, Indonesia, and Japan.

NewsWednesday August 5th, 2020

Understanding the Machine Learning-Based Corona Likelihood Metric…

The government of DKI Jakarta is now using the Corona Likelihood Metric (CLM) to replace the Jakarta Entry and Exit Permit (SIKM). It is reported that the Corona Likelihood Metric is a machine learning-based application of self-assessment.

Self-screening and Self-assessment are the first steps to prevent the spread of COVID19

Self-screening or self-assessment is a method for assessing whether oneself is worthy of being stated in a matter. Self-screening or self-assessment during the COVID19 outbreak helps the government and the community to determine anyone’s high risk of coronavirus exposure. For anyone who is considered at low risk, it is recommended for them to be at home, and for those who are at high risk, it is recommended to take a PCR test.

Honesty is needed in filling the CLM. Lying in answering the questions will affect the machine calculation, and the result won’t be accurate. The CLM can even recommend whether someone should take a PCR test or not, based on the data inputted.

The CLM test can only be done once a week by inputting the identification number. If within a week you have participated in the CLM test, and you are not advised for travel, then you should just stay at home. You may participate in the CLM test again in the following week.

Corona Likelihood Metric will replace SIKM (Jakarta Entry and Exit Permit)

In the future, people who wish to enter Jakarta must take the CLM test that accessible through the Jakarta Kini application (JAKI). The SIKM and CLM have fundamental differences in terms of intended use. 

As reported by Tirto.id, SIKM was used to limit activities in and out of Jakarta during the PSBB, while the CLM used to control community activities so that they feel safe during the transition.

People who wish to enter Jakarta are required to fill in their personal data and health conditions honestly in the CLM. Later, the CLM will process the data and provide a score as a guide whether someone is allowed to enter Jakarta or to just stay at home.

CLM became the first in Indonesia as a machine learning-based self-assessment

Before CLM came into existence, several locals applications such as Halodoc, Grab Health, GoJek Health, Good Doctor, etc. had issued a self-assessment. The difference between the CLM and the existing self-assessment tools is the CLM ability to assess whether a person is eligible to take a PCR test or not.

CLM itself is the first in Indonesia as an independent screening tool that uses the machine learning model. A self-screening tool with machine learning was created in such a way as to be able to measure the likelihood of someone had exposed the COVID19 or not.

The CLM machine reads the history of COVID19 case data from the DKI Jakarta Health Office. At the end of the test, participants will get a score based on answers, ODP/PDP/OTG case status based on Health Office COVID19 case data, and PCR test schedule at the nearest health facilities when approaching a specific score.

How Machine Learning plays a role in building the CLM 

CLM can also be referred to as Machine Learning-Based Clinical Decision Support System (CDSS). CLM is a new breakthrough in the Indonesian IT world in which the implementation is increasing due to the COVID19 outbreak.

To build the CLM first step is taking the preprocessing data, which is cleaning up existing data to then be studied by the machine learning model.

Then, the CLM built several versions of the machine learning model by leveraging 80% of the existing data. There are three criteria used in the testing phase, it is accuracy, precision, and sensitivity. It is the model with the best performance that will be evaluated and be used by the community. 

According to the rapidtest-corona.jakarta.go.id the CLM machine has high accuracy. The CLM engine helps someone in answering questions such as “Am I possibly corona positive?”, “Do I need a test?”, And “What should I do now?”

Machine Learning, reliable in processing giant data with minimum errors

The independent COVID19 risk test application such as the CLM is a step that we should appreciate together. The use of machine learning in the CLM makes it a reliable machine in matters of data processing.

It is different from manual processing data by humans. Machine learning is highly reliable for large data processing, non-stop process, super fast or even close to real-time, and minimum error.

The CLM engine is the result of a collaboration between Jakarta Smart City, DKI Jakarta Health Office, Harvard CLM Team, and KlakklikId. The CLM machine is expected to be able to assist the DKI Jakarta government in distributing PCR test kits to those who really need it.

In addition to self-screening tools such as the CLM, there are also several other AI technology tools that are useful to help the government in controlling the spread of COVID19. For more details, please read our review on The Four Tools With AI Technology To Guard The New Normal.

corona likelihood metric
DX News
Tags

Event
Calendar

Apr
15
2020

Seminar WFH Lebih Mudah dan Efisien dengan…
Sorry, this entry is only available in Indonesian.

Mar
12
2020

Kintone Seminar, 12th March 2020
Pemanfaatan IT Untuk Efesiensi Kerja Integrasi dan Pembuatan Aplikasi tanpa…

Back To Top