DO YOU AGREE OR DISAGREE FROM THIS RESPOND. WHY Can we use unsupervised data min

Question

DO YOU AGREE OR DISAGREE FROM THIS RESPOND. WHY Can we use unsupervised data mining to solve any issues? -The short answer is yes; unsupervised data mining will help to solve issues addressed in the video as this type of data mining allows us to find hidden patterns in data. While supervised data mining gives the best results, in cases like this, it is not always possible to assign a label of interest in problems. As Mr. Stryker said in the future we do not need to know exactly what we are looking for, which is where unsupervised learning can really help us solve the issues he discussed Utilizing unsupervised data mining techniques such as K-Methods can be used to understand how eating habits (i.e. grocery POS data); past treatments; and unintended medicines effect an individual's overall health or the likelihood of someone developing cancer and/or a treatment plan that is the most likely to cure a certain cancer. There are a lot of elements that contribute to disease and Mr. Stryker over simplifies what information needs to be correlated to understand the best way to cure cancer X. There are already tests for genetics to see what medicines work best for certain people, and this could be correlated with medicine data, eating habits data and applied to a person's overall health. This could include many other elements for optimal results. However Mr. Stryker oversimplifies how easily these things can be achieved. We are not there yet, but even in the future with more massive data collected, it is still financially intensive to clean the data and build models for this type of data understanding. This type of information gathering would most likely require distributed data mining, which can cause further complications. While he may have oversimplified the ease and cost effectiveness, yes, it is possible to use unsupervised learning to solve the issues he presented because as he said we basically need these four elements for decisioning: 1) Availability of massive amounts of data; 2) Ability to structure data; 3) Aptitude to discover insights; and 4) Capability to deliver these insights to points of use. With these four elements plus the models I described previously, we can help solve these health correlation issues

Explanation / Answer

Yes, I agree with the response provided about unsupervised machine learning technique used for prediction. As you know two types of machine learning techniques are there which helps us determine the future decision one is supervised and another is unsupervised. Supervised machine learning uses past or historical data and applies algorithms like regression or classification type to predict the future scenario.This type of data mining requires a huge amount of data to train the model and create a robust model to predict an accurate thing for future.But as the paragraph says when we are considering data mining application in healthcare it may be a situation in future where a tremendous data flow will come and handle that data we have to go for distributed data mining technique to handle this volume of data which will be having financial implications.At the same time, unsupervised machine learning classifies the data from its current state and doesnt require any past data.There are lot of techniques available for it like K means clustering, Hierarchial clustering etc. which cluster the similar type of data in to a zone which helps us to identify which category or zone the data belong to.With the ability to do image classification and data classiication unsupervised learning is very popular.It will take less time to learn the data and structure it accordingly

Related Questions

Navigate

• Browse (All)
• Browse D
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.