Hands On: Building A Full Stack Java App (Classical ML)

This course is an element 3 in our sequence, instructing you the best way to construct a full-stack Java utility, from nothing to completely functioning! On this course, we’ll proceed the work from half 2, altering the best way the configuration works from simply uncooked .json information within the file system, to a full config web page on the frontend. This entails utilizing html types, sending and dealing with advanced information buildings to the backend, and saving these information right into a database. We’re additionally introducing extra TypeScript, so we shall be creating TypeScript sorts to make sure the info within the type is shaped appropriately.

We are going to then additionally deal with Machine Studying. We are going to go over what it’s, how we use it on this challenge, and the best way to implement it your self. The stream of the ML within the course is as follows:

1. GridLog reads uncooked HTTP logs from the host

2. GridLog saves uncooked logs

3. GridLog reads uncooked logs from DB and parses into searchable columns

4. Whereas saving the parsed logs, if GridLog detects these are HTTP logs, it would run Machine Studying inference on the logs to try to predict if the logs are malicious or benign

5. If malicious, save the DB entry as attainable intrusion try

6. Mark try in Log Viewer

To get the above working, we might want to use free Machine Studying libraries to do supervised coaching on a dataset supplied to you. As soon as educated, we will run inference on any new incoming HTTP logs.

So for this course, you’ll studying the best way to implement all of this into an already working by merely including in a brand new Docker container to your working docker orchestration file (Docker compose in our case)

Supply Code for this code may be discovered on our GitHub web page which is discovered within the sources part of our Introduction lecture.