How long does it take to become expert in machine learning?
You really cannot master absolutely anything in the world. All you can do is be good at it, gradually get better than the rest with discipline and rigorous practice and then try to improve yourself further.
Anyway, so let me answer some basic questions here. Before starting to get any further I would also like to declare that I have a two years of experience in studying and implementing machine learning algorithms and mapping them to real world problems. Most of the time I do so as a part of my personal work from home projects. I am not directly involved into data engineering as I am a software engineer. I might be wrong in certain sections of my analysis and experienced data engineers might differ from my point of view, but the analysis below is entirely based on my personal experience and Continue Reading
Source: quora.com
Why is Machine Learning difficult to understand?
Getting started with anything new is hard, although the extent varies. The learning curve for ML can indeed be steep for many. But an approach that usually works when you try to learn something new is to learn it in layers . In the first pass, try to get a high-level big picture of what problem you are trying to solve, how are you trying to solve it (at a high-level), and how do you measure that the problem is actually getting solved satisfactorily. Then, you delve deeper into each of these components, gaining more intuition of each. Finally, you get into the technicalities of each of the components.
Here, let me give you a high-level picture of the field.
Problem we are trying to solve : Given some data, the goal of machine learning is to find pattern in the data. There are various settings,
Source: quora.com
What’s the easiest way to learn machine learning ?
My recommendation is a little different from others answering this question; I assume you want to become a star at both Machine Learning AND Engineering.
Why do I draw the distinction? Well, there are lots of folks in the market that are great engineers and there are also lots of folks who are great at machine learning, but there is a severe shortage of great Machine Learning Engineers.
Engineers who are great in both fields are basically unicorns and are at least 10x as valuable as someone who is great in just one of the fields. These are the engineers who don’t just work on algorithms or systems all day but instead launch personalization products in the market. These are the types of engineers who are behind the personalization teams at companies such as Amazon, Netflix, LinkedIn and many
Source: quora.com
Is machine learning really difficult?
No. I wrote my first ML program waaay back in 1982, before there was Internet, Google, GPU computing, laptops, cellphones, digital cameras, desktop PCs, heck before there was almost anything remotely resembling what you see in the tech world around you today.
How did I even discover the existence of such a field? Back then, to educate oneself, you read books. Of course, you had to either go to a library, or in my case, a quaint event called a book fair. I attended a large book fair in New Delhi, India’s capital, and picked up this 800 page tome, a fairly massive affair. Why, I don’t know. After all, I was at the Indian Institute of Technology, Kanpur, studying to become an electrical engineer.
Hofstadter’s first book, and in my opinion, still his best, was an utter revelation to me. It opene
Source: quora.com
Is machine learning hard or easy to learn for an average software developer?
If you are asking about the tools, any developer can learn and code to RapidMiner, TensorFlow or Watson. It’s just a new frameworks and set of api’s. If you want to know what is going on inside the tools then buckle-up, its going to be a rough ride.
The science and math involved is very complicated, and that instructs you as to which algorithm of the 40+ available to use to achieve your result. No part of this understanding is simple math, each of the algorithms is tested against each other to see which achieves the best result on a given sample set. There is no simple “if this then that” paradigm here, it is an art.
The other hurdle is the data itself. Most ML processes must impute missing data from the input in order for the algorithm to work. So if a person leaves a question blank, what s
Source: quora.com
Is machine learning hard to master?
Any one field is hard to master and generally takes around 10,000 hours to be considered an “expert”, but the key here is that you don’t really need to be an expert to be able to use machine learning.
Here are some of the basics you’ll need for machine learning:
basic handle of linear algebra. Think about this, the linear regression formula you learned in high school Algebra (y=mx+b), is one of the most basic forms of machine learning. You can give some examples of x and y and come up with values of m and b that model the y variable.
a basic understanding of some standard machine learning algorithms, how they are classified, and how each one can be optimized and monitored for maximal performance: that is, a set of unsupervised and supervised algorithms. A good place to start would be this web
Source: quora.com
Difficulty 1: Adapting to the Problem Domain
How many mathematicians study Linguistics? How many mathematicians study Healthcare? So why are we any good at solving problems in these fields?
The art of being a Mathematician comes from the ability to abstract a problem in a manner that makes it solvable. In Linguistics, we can treat each “phone” as a discrete variable and create a model that determines the joint distribution between each phone. In Healthcare, we can build a model that picks up latent features in X-rays that discern a disease.
Source: medium.com
Difficulty 2: Identifying and Ignoring the Noise
Noise is second to none in statistics, machine learning and data science. Honestly, it’s everywhere. From dirty data, to rogue data points, to literature built on weak foundations, to models capturing latent bias: noise is literally everywhere.
Machine Learning models generally perform by minimising the squared sum of errors (or some form of misclassification measure) but when you’re researching a new topic or getting feedback from a colleague, noise can be pretty hard to define the last thing you want to do is be chasing down the rabbit hole.
There are a few ways to get around it:
Speak to reliable people often, keep them close
Learn how to spot nonsense, keep it at a distance
Fail often, fail quick.
Experiment more, speak to people more, try more things and eventually you’ll begin to recognise and ‘smell’ noise. You’ll avert it, and progress quicker.
As an example: many algorithms have a high accuracy rating because the dependant variable happens so infrequently. E.g. a model which predicts how many people in London get struck by lightening on a daily basis will almost certainly be 99.9999% correct without any training. The “noise” is recognising that people don’t get struck by lightening that often, and by adjusting your model for it.
Source: medium.com
Difficulty 3: Getting Good Education
Markus Leo on Unsplash Education is so important in this field because the domain of knowledge required is so broad. From computer science, to maths, to algorithms, to statistics: there’s a lot to cover in a relatively short amount of time.
Formal education (like University) is one thing but education in machine learning really surpasses that. Practitioners have to develop an ability to quickly learn things themselves and be able to implement them well.
The reason why this is so important (and so difficult) is that it’s tempting at times to find a github repository where someone else has spent some time solving the same problem you have, pulling their code and applying it to your problem. The solution make look ok but plenty of things can just get missed in between all of it and there’s no comparison to having the fundamental understanding.
Source: medium.com
Difficulty 4: Publishing Negative Results
Negative results happen all the time, they’re hard, but they happen. You have to recognise that negative results are also results and that they should be welcomed.
Machine Learning has two sides to it: the theoretical and the applied side. Theorists will publish less frequently with the hope of making a bigger splash and applied academics will tend to publish more often but solve bigger problems.
However in the pursuit of experimentation or in the pursuit of publishing, a lot of negative results are often put to the side and not overly discussed. This then leads to other practitioners repeating these same experiments and at the aggregate, a lot of time is wasted. This inefficiency also breeds a form of ego where people are respected by only the ‘positive’ results they’ve discovered, rather than the results they can confirm to be simply incomplete.
Everyone benefits if we can classify problems better.
Francisco Moreno
Source: medium.com