Q1: What is the difference between data mining and machine learning?

Ans: Data mining is considered the process of extracting useful information from a vast amount of data. It is a tool we use it when we are looking for miningful and relevant information. On the other hand, Machine learning is the design, study, and development of algorithms that permit machines to learn without human intervention. In other words, it is a tool we use to make machines smarter.

Q2: Can AI application in Natural Language Processing already take care of discourse overlapping now in dialogue machine translation? Or in multilingual (more than bilingual) discourse?

Ans: There are already AI products such as Siri or Google Voice Assistant, which can help us with multilingual translation. However, at the same time multilingual dialogue translation technology is not yet mature.

Q3: What if the data that the AI gathered was only half true or limited by perspective from the beginning, but unknown at the time, what is the next step?

Ans: There are currently two types of deep learning, one is supervised learning, and the other is unsupervised learning. Unsupervised learning can start training when the data is not clearly defined.

Q4: As it said in the lesson, it need more sensor detection to public and AI can aid in detection to corona-virus,. without any input, the AI can't do anything for telling people whether they infected corona-virus or not

Ans: Right. Without any input, AI can not help us analyze people who infected or not. But now we have some information about people like ages, gender, symptom etc. also, we can predict the virus structure more precisely which may be appear in the future by using AI to analyze the DNA sequence.

Q5: What is the accuracy of  AI Predictions on COVID-19? 

Ans: AI help people confirm their movements and check the contact. According to these information, they can know the virus tracking and take some action immediately.

Q6: How do you think the data can be used by Government agencies for their decision making processes - e.g. when schools needed to be closed. In your opinion, what might be the level of barriers from ethical, political or data reliability perspectives?

Ans: There is no doubt that any government decision must be based on scientific data. I believe that everything must be based on science.

Q7: Is it possible to use Deep learning to predict ahead of virus mutation to prevent future virus like corona and such? What percentage of accuracy from the result that you will considered practical?

Ans: All of this is possible, but we can't guarantee the completion of it, because there are still many unknowns to us. The availability and accuracy of data is also very important and must also be verified first. aat this stage it is still to early to provide you with the correct accuracy value.

Q8: The label we used, is it from the dataset ? and if we want to make our own labeled, how to make it prof ?

Ans: Yes, we often use the label from the datasets. If you want to make your own label, it must be based on the original label and follow the certain rules. If you can convince others your rule of new label is reasonable, it’s ok.

Q9: When we talk about Big data analysis, how can we know how big data we need to use?

Ans: It depends on the topic you use and in this field, how big data is enough to proof the effectiveness of the result so as to let your study able to apply in the life.

Q10: How can we evaluate the leakage or overkill of the model?

Ans: Leakage means, the model learns from data that wouldn’t (or shouldn’t) be available in a real-world scenario. In other words, when the data you are using to train a machine learning algorithm happens to have the information you are trying to predict. Overkill on the other hand, is the ability to process large numbers of features makes deep learning very powerful when dealing with unstructured data. Deep learning algorithms can be overkill for less complex problems because they require access to a vast amount of data to be effective.

Q11: Can AI predict the Lotto Number?

Ans: AI doesn't understand causation, merely correlation, which is often good enough to make a good guess. The lotto is designed to be random. It is not correlated with anything. Thus, AI cannot be used to make any meaningful guesses.

Q12: Is it possible to make AI plays forex by its own and the owner just get the income from it? 

Ans: Yes, it is possible. It benefits forex traders on many levels. It analyzes massive amounts of data and uses current stats and trends to provide better market forecasts. More importantly, it automates key parts of the trading process and lets you track your performance in real-time, helping you identify core problems and fix them instantly.

Q13: How AI will help cope with the climate change?

Ans: AI can help us cope with climate change in ways such as: Monitor agricultural emissions and deforestation, deployment of autonomous vehicles, automating the analysis of images of power plants to get regular updates on emissions, between others.

Q14: How about the hardware specifications such as camera to get the good data enough for AI analyzing?

Ans: It depends on the approach to the problem. Hardware is no longer important in the current urgent situation. Available resources should be used to localize and prevent infection. Can use natural language recognition processing to detect keywords when talking in the phone so that the area can appear.

Q15: How can AI teach the human ethics to the AI doctors?

Ans: AI system has explicit ethical agents which are programmed to calculate what is ethical on their own by applying ethical principles to complex situations such as diagnosing depression in a patient and then prescribing treatment as well as monitoring the patients’ symptoms. This means that the system will have to have the ability to make decisions and select courses of action which are conversant with the codes of ethics of its practitioners. Professional ethics and machine ethics must thus operate in a way that benefits of the patient.

Q16: AI ahs been developed back many years, but why with the COVID19 pandemic, the process of developing vaccine or any methodology to cure the virus has take long period of time?

Ans: The process developing vaccine has take long period a long period time  and its effectiveness may still need to be tested before approval and distribution. Even if proven effective, the vaccine might not be widely available. AI is faster but still takes time, AI is not used to cure the virus itself. However,  AI can support researchers to speed up the process.

REF: TTAIC & AIGO & III