Cloud Machine Learning
-
HenryIdomo
Cloud Machine Learning
Using machine learning, what is the difference between saying "I have used machine learning to build up a machine learning model" and "I have used machine learning to determine a heuristic ie dividing the two sides is some boolean function, ex. x > 5"?
Re: Cloud Machine Learning
I'm no expert in machine learning. In fact far from it. Since starting the project last year I've discontinued after running into the problem of not a large data set. For my model there is simply no database out there that I'm aware of that exists and is properly labelled for a Single Shot Multibox Detector (SSD).
Now answering your question. I would have to say that generally in the machine learning industry you are "using machine learning to build up a machine learning model", which in my view means that you are taking methods and functions from ai research, and using it to create a mathematical model varying in degrees of complexity that will given an input produces an output, right?
Now for using ai to determine a heuristic is a possible application of machine learning, however, in some cases, speed is not the most sought after, more accuracy. For example, if designing a machine learning model that takes in two fingerprints, one test subject, one base subject (most likely from a database). A heuristic by definition (according to wikipedia) is 'a technique designed for solving a problem more quickly when classic methods are too slow' and 'this is achieved by trading optimality, completeness, accuracy, or precision for speed'. In the case of the fingerprint example, accuracy is key, not speed, seeing as results could possible result in legal ramifications against someone (guilty or innocent).
Again, I'm no expert in the ever expanding world of machine learning, so take this as a grain of salt, and make sure to research online.
Hope this helps,
Ben10
Now answering your question. I would have to say that generally in the machine learning industry you are "using machine learning to build up a machine learning model", which in my view means that you are taking methods and functions from ai research, and using it to create a mathematical model varying in degrees of complexity that will given an input produces an output, right?
Now for using ai to determine a heuristic is a possible application of machine learning, however, in some cases, speed is not the most sought after, more accuracy. For example, if designing a machine learning model that takes in two fingerprints, one test subject, one base subject (most likely from a database). A heuristic by definition (according to wikipedia) is 'a technique designed for solving a problem more quickly when classic methods are too slow' and 'this is achieved by trading optimality, completeness, accuracy, or precision for speed'. In the case of the fingerprint example, accuracy is key, not speed, seeing as results could possible result in legal ramifications against someone (guilty or innocent).
Again, I'm no expert in the ever expanding world of machine learning, so take this as a grain of salt, and make sure to research online.
Hope this helps,
Ben10
Re: Cloud Machine Learning
One year later and I've found a free solution.
Google Colab https://colab.research.google.com is a free jupyter notebook environment that runs only on the cloud.
In doing so you can run jupyter notebooks like https://colab.research.google.com/githu ... tion.ipynb
Which preforms object detection.
For free you get access to run Python code on:
GPU: 1xTesla K80 , compute 3.7, having 2496 CUDA cores , 12GB GDDR5 VRAM
CPU: 1xsingle core hyper threaded Xeon Processors @2.3Ghz i.e(1 core, 2 threads)
RAM: ~12.6 GB Available
Disk: ~33 GB Available
Google Colab https://colab.research.google.com is a free jupyter notebook environment that runs only on the cloud.
In doing so you can run jupyter notebooks like https://colab.research.google.com/githu ... tion.ipynb
Which preforms object detection.
For free you get access to run Python code on:
GPU: 1xTesla K80 , compute 3.7, having 2496 CUDA cores , 12GB GDDR5 VRAM
CPU: 1xsingle core hyper threaded Xeon Processors @2.3Ghz i.e(1 core, 2 threads)
RAM: ~12.6 GB Available
Disk: ~33 GB Available