Are you interested in Machine Learning? Unsure if you should invest the time into learning more about it?Well, here are the key steps in any Machine Learning workflow:<ul><li>The Dreaded Dataset</li><li>Data Preprocessing from Hell</li><li>Finally Develop the Model</li><li>Give the Model a Workout</li><li>Evaluate your Unfortunate Results</li><li>Do it all over again. But better.</li></ul>Sorry, there aren’t nine. But the joke will make sense later!

The Nine Circles of Machine Learning

I lied, there’s one more step. Before anything else, you must define the problem your machine learning algorithm is going to solve. Or try to solve.Is it related to text, images, video? Sound? You can mix and match, too.And is it classification, segmentation? Detection?This will inform the dataset you’re looking for or — God forbid — building. And also which type of model you should use.

Define the Problem

Here’s where to look for a dataset:<ul><li>Dataset search engines: kaggle and google dataset search.</li><li>Papers with similar problems: google scholar and papers with code (pwc is a surprise tool that will help us later).</li></ul>If kaggle shrugged and google laughed at you, if none of the researchers made their datasets available, or if you’re a trailblazer thinking of using data no one has gathered before… Time to get your hands dirty. Set up some crawlers, scour google and youtube. And may the force be with you.

The Dreaded Dataset

Preprocessing involves preparing the data, making sure it’s in the format the model requires.If you’re working with images, always normalise!! This isn’t the first or the most important step, but it’s the one that always gets me. And use open CV to turn images into arrays.If you’re not, I can’t help you here. Google and stack overflow will guide you.

Data Preprocessing from Hell

Congratulations! Like Dante, you have muddled through the 9 rings of eternal damnation and survived to tell the tale. These are the pearly gates, this is the fun part.Someone said once that there’s nothing new under the Sun, I think. And they were right. The first thing you should do is try and replicate models from papers with code. See, I told you it would help us.Papers with code kind enough to sort papers by performance and provide github links. All you have to do is clone the repository and get the model working, try it out with your dataset, and adjust it as needed.

Finally Develop the Model

It’s time to train the model. Hopefully you set data aside for testing. My bad, I should have mentioned that earlier.Usually we do 70% of the data for training, and 30% for testing. You can also have 20% for testing and 10% as a hold out.Training can happen in seconds, or it can take days. It all depends on how complex your data and your model are.This is the part where you sit back and relax. Grab some coffee, take a nap.

Give the Model a Workout

Hush, now! It’s okay. If it were perfect that would mean there’s no more fun to be had!As far as metrics go, I like to go all out and catch them all. Keep in mind that if you have an unbalanced dataset (where one class has much less data) the accuracy will be misleading. Look at the sensitivity, specificity, all of those.

Evaluate your Unfortunate Results

Check the training metrics.Are they much better than the test metrics? That means the model learned a bit too well. It memorised the data and doesn’t perform as well on new information. You need more data, or a different model altogether.Now if the training metrics suck? That means the model sucked. You need to tweak it. Add more epochs or layers, change the loss or activation function. If nothing works, it might be time to switch out the model.This is called overfitting and underfitting, if you want to be fancy.

Do it all over again. But better.

All jokes aside, the joyous world of Machine Learning is well within your reach. It’s not as hard as it seems, and it is just as fun as I made it seem. Pinky promise.And you’re sure to have a large community of people banging their heads over the same exact problems you’re having.

Go forth and experiment!

This is as much for me as it is for you. I love ranting about my job.

computerscience

Computer Science

artificialintelligence

Artificial Intelligence

So you want to teach computers

Bullet journals and planners aren’t for everyone. But everyone should have a way to keep themselves accountable to their health and projects.The benefit of setting up a system on your phone is that the upkeep requires very little effort, because there software tools in place to do the work for you. Simply have the patience to set up the system, and reap the benefits.

A Holistic Health and Productivity System for your Phone

The Health app is powerful. It can be integrated with a other applications and keep all your information in one place.Any water, nutrition, exercise, or sleep apps you use can and should be synced automatically with Health.

The Health App

This is the key to a useable system. The more of a hassle it is to input data, the less likely you are to keep it up.So use widgets and shortcuts! When looking for a good app, also test that they have useful widgets and shortcuts available.A water tracking app, for example, should ideally have:<ul><li>A widget with your goal and how close you are to reaching it;</li><li>A way to automatically add a certain amount of water, through shortcuts or a widget.</li></ul>Same for any habit tracker or nutrition app

Simple input methods!

It’s important not to have any superfluous apps in your system. If it’s not useful, it’s wasting your time and you’re less likely to update it.Think about your routine and your needs, and pick your apps accordingly. Here are some you might find useful:<ul><li>Water;</li><li>Nutrition;</li><li>Exercise;</li><li>Sleep;</li><li>Tasks;</li><li>Calendar.</li></ul>

What apps do you need?

Some aspects of the default apps are inconvenient. Adding water or nutrition data, for example, takes diving deep into the recesses of the Health app.But sometimes they’re fine! Better than fine, sometimes they’re well integrated and have all the features you need. It’s worth it to check first, before you pay for a more fancy app you might not need.

Is the default app enough?

Habit will not disappoint. It’s feature complete and has good widgets. The shortcuts are lacking, but it is free. It also integrates with the Health app. If you want to track sleep, steps, exercise, water, meditation… The sky is the limit, and Habit is your rocket.

So you need a habit tracker…

I’ve got you. Waterful has a friendly octopus that needs you to drink water to survive. It has useful widgets, and you can also configure shortcuts to add any amount of water you want.Remember to untick the show on run option on the shortcut and you’re one click away from logging your water.

So you need a water tracker…

The default is fine. It integrates with the sleep focus mode, and the paid ones don’t add much value.

So you need a sleep tracking app

Avoid subscription based apps, those will build up quickly and aren’t usually that special. The default reminders app is fine, but if you need a bit more Things 3 is great. It has useful widgets, of course, as well as shortcuts. It also integrates with your calendar, which is useful to better visualise your day.Todoist and TickTick are fine, but some of the best features are behind the subscription pay wall.

So you need a to-do list app…

The apple calendar is gorgeous and plenty functional. You can your google and microsoft calendars, and add events directly through it.

So you need a calendar app…

I’ve got nothing. Sorry. I could sit here and recommend you an app I don’t use, but my integrity won’t allow it. And a nutrition app is just not something I care about and I know I wouldn’t update it.Here’s an important lesson. Know what you don’t need. It’s often even better than knowing what you need. You can quote me on that.

So you need a nutrition app…

All the apps I recommend integrate with it. You can use their complications for easy access. Honestly, it becomes so easy to keep up the system even I do it.Also use the shortcuts app, you can add most of your shortcuts to the watch. I have my water shortcut and a few others to turn lights on and off.

If you have an Apple Watch:

<ul><li>Tide: I’m a hypocrite. This meditation app has a subscription. But this is worth it for me. Tide makes me meditate everyday, and it syncs with the Health app. Find your Tide, it doesn’t have to be paid.</li><li>Drafts: Useful for quick capture of notes, on your phone or apple watch. The pro version is subscription based unfortunately, and I’m looking for alternatives. Please do tell me if you know any.</li></ul>

Honourable mentions! Part 1

<ul><li>EMMO: If you look past the weird translations, this is wonderful. There are widgets for quick capture of you emotions. You can add pictures, stickers, timestamps. Just don’t try to delete any of your emotions because it breaks everything.</li><li>Stoic: Good to guide your day. It asks you a few quick questions in the morning and evening. Personally, I use Stoic in the morning and EMMO in the evening. I like that Stoic guides you, it’s just what I need when I wake up.</li></ul>

Honourable mentions! Part 2

If for some reason you would like to know what is going on in the world, you poor soul, I recommend downloading all the news apps you like and putting their notifications in scheduled summaries.I have mine during breakfast and the ride home from work, and I read all the headlines and click the articles I want to know more about.If you don’t want to know, I feel you. If it’s really important I’m sure you’ll eventually hear about it.

Honourable mentions! News apps

Set up your system to be deeply integrated and have simple input methods. Shortcuts and widgets are your friends.Don’t add apps you don’t need, they’re just clutter. When looking for the right app, make sure it has all the features you need, widgets and shortcuts, and that it integrates well into your system.Now go forth and be productive!

TL;DR

I’ve always had a productivity system in place, but this one is the most convenient. Don’t underestimate your phone. It’s not just good for instagram.

So you want to teach computers

Make your phone productive!

Morgan G