The short-hand answer is, theoretically, yes.

In practice, it is much more difficult to answer. There needs to be a reliable dataset for training, which is very difficult to achieve. Why? Because there is at present a vast amount of misinformation being disseminated. For example, in the U.S. when a patient is admitted, if he or she dies of heart disease, but tested positive for the COVID-19 virus, the administration forces the hospital staff to sign off as a coronavirus death, even if the patient was already a heart disease victim and the COVID-19 virus was 'inert'.

Misleading or outright incorrect data is going to muddle research results on COVID-19 for years I'm afraid. Statistics is the only branch of mathematics which can be politicized and "spun".

However, once the dust has settled and real researchers can acquire accurate data with all relevant factors (age, overall health, etc) accounted for, the lessons learned will help the world to better respond to the next outbreak. There is SO MUCH data being assembled right now with the current pandemic, but unfortunately much of it is sensationalized to sell papers and keep the public scared. Once the real information is collected then machine learning algorithms and other AI tools will help us to analyze the process and find an effective strategy for response and treatment.

Yes, it is possible but it will require some time to achieve that

Yes it possible!

yes it is possible but enough time is required to obtain enough and robust data

Yes, very possible

It is already happening In four large companies that have openend their platforms. You can find the details in my article https://aqalgroup.us13.list-manage.com/track/click?u=63a10496759d992f7198eb4d7&id=2ac1df1cf8&e=8d53cec068

The matching of pathology and pharmacology may be able to screen some candidate drugs. This process machine learning will help.

The "already happening" answers are not accompanied by evidence.

The Daily Mail article talks about using game computers' processing power to "understand how viral proteins work and how therapies can be designed to stop them".

Mariana's article says "They have opened their platforms to allow scientists around the world to leverage their collective intelligence to expedite the development of an antivirus and a vaccine."

Both articles are about allocating computing resources to the problem, which is different from saying that machine learning is helping or will help.

That's why I'm only recommending Cheggang's answer above, which at least directly addresses the question and correctly points to a potential use of machine learning to help find treatment.

Yes, and not just in this particular case. Using machine learning to analyse genome is well established also for more complex organisms than a virus. Besides, machine learning plays an important role in finding patterns why certain drugs/medication works for patients while not for others while also accounting for influences of environment and life style. Predicting when, what and how good machine learning leads to COVID-19 medication is however not possible.

You may want to browse here too https://spectrum.ieee.org/the-human-os/artificial-intelligence/medical-ai/companies-ai-coronavirus

Can the machine learning figure out your researchgate password? If yes, then it's a promising solution 😊

I think, it is possible.. But don't how to get there....

Deep learning can help in finding a medicine for COVID-19

It can definitely help and i believe it's already happening. the algorithms needs to be fed data on specific drugs that have been tested and their results and it might be able to find correlations that we previously missed. More data gives better results and with the increase number of cases and test i believe machine learning is a very promising path.

Thank you all.

I think ML/DL can help. Furthermore, with proper design of the algorithms, ML/DL could provide more information such as the importances (rankings) of the medicine and their correlations with COVID-19.

Thank you Po-Ya Hsu.

Yes it is but first relevant data is important in hands.

Kayode Owa

In my opinion, I think yes but need a real dataset.

Yes it can help. the primary goal is not only to forecast accurately, but to find factors that impact transmission rate of COVID-19. If you find variables that look like they impact the transmission rate, it will help a lot.

Programming a fed algorithm:

The disease is classified according to its structure (DNA):

1. COVID-19 previous.

2. COVID-19 current.

3. COVID-19 renewable (potential).

The affected areas (countries) are divided by the type of disease.

America - Current

- China - Previous

Middle East - Present

Identify known and potential disease pathways within the probability algorithm.

Yes, of course. I think that since there are different datasets available online, you can test different types of algorithms for both regression and classification problems.

Hello Prof. Dr. Chandrakant,

Yes, as per my study, the self-learning approach(Reinforcement Learning) and generative approach(Generative Adversarial Network) together can be used for finding the medicine for COVID19. But the molecular structure generated using GAN should be tested accurately so that the RL algorithm can learn the molecular structure more accurately. I think there is a gap between the testing of the drug generated by GAN and the learning mechanism of RL.

The short-hand answer is, theoretically, yes.

In practice, it is much more difficult to answer. There needs to be a reliable dataset for training, which is very difficult to achieve. Why? Because there is at present a vast amount of misinformation being disseminated. For example, in the U.S. when a patient is admitted, if he or she dies of heart disease, but tested positive for the COVID-19 virus, the administration forces the hospital staff to sign off as a coronavirus death, even if the patient was already a heart disease victim and the COVID-19 virus was 'inert'.

Misleading or outright incorrect data is going to muddle research results on COVID-19 for years I'm afraid. Statistics is the only branch of mathematics which can be politicized and "spun".

However, once the dust has settled and real researchers can acquire accurate data with all relevant factors (age, overall health, etc) accounted for, the lessons learned will help the world to better respond to the next outbreak. There is SO MUCH data being assembled right now with the current pandemic, but unfortunately much of it is sensationalized to sell papers and keep the public scared. Once the real information is collected then machine learning algorithms and other AI tools will help us to analyze the process and find an effective strategy for response and treatment.

This question is a very interesting topic of research in the field of Bioinformatics as this has so much to do with drug-human interactions.

Yes, ML, AI and Data Science can be used to better understand the COVID-19 epidemic. What types of data do you want to use? if it's a text document then you can analyze the sentiment of data for better understanding (using the NPL process). if it is image data then you can predict COVID-19 positive or negative If you have enough data.

For some days, Kaggle starts the UNCOVER COVID-19 Challenge: https://www.kaggle.com/allen-institute-for-ai/CORD-19-research-challenge/tasks. Everyone can participate in this Challenge. They are providing a huge amount of data for the researchers.

For some days, I am working image classification with Tensorflow by implementing deep learning models. I am training COVID-19 patients Xray and CT Scan images for prediction.

It is possible, Doctors and Computers Programmers just need to work together. Its like creating an AI with a hologram. Hologram that displays humans internal structure, and what happens internally when it is harmed, and how it tries to recover. Imagine having a virtual human, one that you can make sick by punching code, and then sit back and observe what really changes in the blood or organs when affected by that sickness. Recording the changes to show how the virtual body tries healing itself. A database can provide medical information to the AI on how illnesses are treated. If data is collected well, Doctors and Al can easily tell how to fix the virtual body, when it is affected with Covid 19. It is time for Doctors and Computer Scientists to work together. With deep learning and an expert doctor in the shape of a software, maybe Covid 19 would have never happened. Computers are more efficient, so they can be of huge help to the medical field.

Yes, Deep learning can help if we get a real data about the recovered people and their medical report.

Yes, It can help if you have the data detail about the COVID-19 you can do classification. So, Need to real dataset about COVID-19.

yes, I think it is possible to find medicine for COVID-19 by using neural networks or artificial intelligence methods.

You may follow our survey at https://www.preprints.org/manuscript/202004.0383/v1 and Preprint Artificial Intelligence (AI) and Big Data for Coronavirus (C...

Besides, we provide a set of challenges and recommendations for the research community.

Any comment, suggestions, and citation are welcome and much appreciated.

Indeed it will help within its own limit of performance : https://hbr.org/2020/04/leveraging-ai-to-battle-this-pandemic-and-the-next-one

Dear Dr.Chandrakant Naikodi

Hope these articles may help you to find your answer.

https://www.geekwire.com/2020/ai-helping-scientists-fight-covid-19-robots-predicting-future/

https://www.datanami.com/2020/03/23/can-ai-find-a-cure-for-covid-19/

https://towardsdatascience.com/how-can-ai-help-with-the-covid-19-vaccine-search-a68d40fc0cb0

Regards

This is obviously possible, only if one has enough data sample with an ample number of independent variables or features.

Thank you all.

I recommend you this survey

https://www.preprints.org/manuscript/202004.0383/v1

It depends on the data set that will be given to be used in ML/DL.

It can have a role in corona vaccination developmental analysis for example see : https://www.wired.com/story/opinion-ai-powered-biotech-can-help-deploy-a-vaccine-in-record-time/

Article Artificial Intelligence (AI) applications for COVID-19 pandemic

https://www.drugtargetreview.com/news/59650/ai-and-cloud-computing-used-to-develop-covid-19-vaccine/

https://www.healio.com/news/primary-care/20200511/artificial-intelligence-transforms-search-for-covid19-vaccines-cure

Dear Dr.Chandrakant Naikodi,

The answer for your query is absolutely Yes.

But in the current situation it is not possible. It requires more time to gather the perfect data and more number of samples to investigate it through ML/DL. Because if more number of samples are considered for training, which it returns accurate result. So it needs to wait for the collection of proper samples from a source. Hope it may help to clarify you.

Thanks

I think many companies and institutes have already applied ML, DL in their attempts to discover a treatment for COVID-19.

Yes, artificial intelligence, machine learning and other Industry 4.0 technologies are used to improve analytical processes in which large data sets are multi-critically processed, e.g. in the field of conducting research on the improvement of medical therapies, including the development of a vaccine, and to develop projections of the future development of the Coronavirus pandemic SARS-CoV-2 (Covid-19).

Best regards,

Dariusz Prokopowicz

I think the answer is Yes it is possible if one can get a reliable dataset with enough entries.

DL/ML is fully dependent on the dataset. so if you have that you can investigate whatever you want.