Another good news about that is that a lot of effort is being put into what is called democratization of AI. advertisement Continue reading below Therefore, it depends on how custom the use case is. If your use case is common, you'll probably find something that doesn't even require you to write code to use it. Therefore, all top cloud providers have a tool called AutoML. Google, Microsoft, and Amazon have the tools to serve datasets. There are several different ways. You can use an existing pre-trained dataset to solve the problem.
There are also black box ghost mannequin effect tools such as Monkey Learn and BigML that you can use out of the box. And you say, "OK, this is my problem. I'll do it through the tool. Please tell me the predictions you need." And the prediction can be anything, such as images, classification of numbers, and so on. It depends on how customized your use case is. And there are different layers in between. I love being able to customize it as I can come up with more innovative solutions away from common use cases. advertisement Continue reading below So my approach requires a little more coding than usual, but nonetheless, in some of the examples I've used, it doesn't require any coding at all.
If you use an Uber tool called Ludwig, you only need to provide a configuration file and you don't have to learn coding. However, the more unique the use case, the more specific and novel you want to be about a particular solution, the more knowledge you need from a scripting perspective and the AI ​​knowledge needed to do so. What are some of the foundations that people interested in this should learn? HB: I think one of the basics is called ETL. This means "extract, convert, read". These are the principles that always apply. There is something called a pipeline in machine learning. Therefore, regardless of the mode, platform, or technology you use, you need to prepare your data to make predictions. advertisement Continue reading below Whatever problem you have, you pull data from Google Analytics or Search. Download in various formats. Machine learning model.