After looking around the forum and the various categories available, I can’t find an answer to a basic question: to what practical use can I, as a retired EE, put Ml to in my life, home, or hobbies?
I KNOW what ML is, I’ve got several book references on learning ML, I understand to a limit how ML works, I can code in python (and C++ if need be), I’m comfortable with RPi’s Arduinos and ESP-32 micro-controllers, and all that … I more or less get all the WHAT and HOW.
I just can’t see the WHY I’d want to learn ML as an individual who no longer works for a living, other than as another hobby/interest. Other than making another little robot with vision or another personal assistant which reads me the news and the weather, what day-to-day practical application would motivate me to participate in this growing field?
I’m not trolling, or trying to start an argument, I’m just trying to find a reason why I should invest the time, at the expense of my many other hobbies and interests. I WANT to learn this stuff, but there HAS to be a purpose to the effort, aside from just learning something new.
I know it sounds ridiculous, that a techie is struggling with this new tool, but I could use some prodding here. Can anyone tell me what got them motivated to actually learn this stuff and start an ML project, other than as an IT assignment for work, or to build a cute little SBC robot that follows you around the house? What did you use your ML learning to do?
Interesting question, can share some of the details of products and areas you have worked in or are interested in from a hobbyiest point of view?
If there is a pattern you can recognise but cannot easily build a thresholding rule based system to detect what you want to alert for that could be something to build a project for. e.g. farming sounds, gases, disease patterns air quality, semaphore hand signals, etc. More commercial would be Predictive maintenance based on audio or other wear indicating signals.
My interests are in community outreach projects and healthcare, I also hobby in car maintenance. Some examples of projects I’ve worked on centre around those topics. Particularly as a hobbyiest you could solve some fairly niche problems you or friends / family could be be having that don’t have a commercial value.
We have an expert community that creates projects that are shared in our docs you may have missed:
Happy to have a call with you if you wanted to discuss your past work and projects for inspiration.
Hi Eoin,
Greatly appreciate your response, I was starting to think I wouldn’t get a single reply. I know my question is somewhat bizarre inserted into an ML enthusiast’s forum, but its got me stalled.
My interests and hobbies run the gamut : astronomy, astro- and general photography, computer tech, Christmas LED displays, home automation, electronics, running and cycling, 3D printing, and so on. I worked for 30 years as an inside sales and technical specialist for an outfit that sells process controls (pig iron and computers/software) into central and western Canada - retired in Dec09.
Something to get me excited and motivated enough to dive into with ML doesn’t have to be commercial, but I know myself: if a project doesn’t have some end-use purpose for me aside from learning something new, I’ll get distracted by the next shiny thing I see that DOES offer a purpose, and I won’t finish my ML adventure.
I don’t have anything against learning for learning’s sake, many of my hobbies and interests fall into that category. But I’m already over-subscribed for 30hr/day to my existing interests, so a new one comes in at the expense of something else.
Anyway … I’ll take a look at that document you linked. I thought I’d already hit all the low hanging fruit in my exploratory search, but I might have missed that one. Thanks again, Eoin!
Oh thats and interesting background, I would love to hear thoughts on projects for industrial predictive maintenance, I have seen a number of printer motor demos at events, but again these are not very useful for you at home when a motor costs far less than the effort to build, unless you have a well stocked freezer with a compressor motor
You have a a nice set of hobbies to explore! I have played around with different USB cameras and a PI, if you have a USB telescope you could train a model to detect patterns or clusters of stars etc that you could train a model to detect (I’m clearly not into this as a hobby yet)
