AI is transforming the way businesses operate, and nearly every company is exploring how to leverage this technology.
As a result, the demand for AI and machine learning skills has skyrocketed in recent years.
With nearly four years of experience in AI/ML, I’ve decided to create the ultimate guide to help you enter this rapidly growing field.
Why work in AI/ML?
It’s no secret that AI and machine learning are some of the most desired technologies nowadays.
Being well-versed in these fields will open many career opportunities going forward, not to mention that you will be at the forefront of scientific advancement.
And to be blunt, you will be paid a lot.
According to Levelsfyi, the median salary for a machine learning engineer is £93k, and for an AI engineer is £75k. Whereas for a data scientist, it is £70k, and software engineer is £83k.
Don’t get me wrong; these are super high salaries on their own, but AI/ML will give you that edge, and the difference will likely grow more prominent in the future.
You also don’t need a PhD in computer science, maths, or physics to work on AI/ML. Good engineering and problem-solving skills, along with a good understanding of the fundamental ML concepts, are enough.
Most jobs are not research jobs but more implementing AI/ML solutions to real-life problems.
For example, I work as a machine learning engineer, but I don’t do research. I aim to use algorithms and apply them to business problems to benefit the customers and, thus, the company.
Below are jobs that use AI/ML:
- Machine Learning Engineer
- AI Engineer
- Research Scientist
- Research Engineer
- Data Scientist
- Software Engineer (AI/ML focus)
- Data Engineer (AI/ML focus)
- Machine Learning Platform Engineer
- Applied Scientist
They all have different requirements and skills, so there will be something that suits you well.
If you want to learn more about the roles above, I recommend reading some of my previous articles.
Should You Become A Data Scientist, Data Analyst Or Data Engineer?
Explaining the differences and requirements between the various data rolesmedium.com
Right, let’s now get into the roadmap!
Maths
I’d argue that solid mathematics skills are probably the most essential for any tech professional, especially if you are working with AI/ML.
You need a good grounding to understand how AI and ML models work under the hood. This will help you better debug them and develop intuition about how to work with them.
Don’t get me wrong; you don’t need a PhD in quantum physics, but you should be knowledgeable in the following three areas.
- Linear Algebra — to understand how matrices, eigenvalues and vectors work, which are used everywhere in AI and machine learning.
- Calculus — to understand how AI actually learns using algorithms like gradient descent and backpropagation that utilise differentiation and integration.
- Statistics — to understand the probabilistic nature of machine learning models through learning probability distributions, statistical inference and Bayesian statistics.
Resources:
This is pretty much all you need; if anything, it’s slightly overkill in some aspects!
Timeline: Depending on background, this should take you a couple/few months to get up to speed.
I have in-depth breakdowns of the maths you need for Data Science, which is equally applicable here for AI/ML.
Python
Python is the gold standard and the go-to programming language for machine learning and AI.
Beginners often get caught up in the so-called “best way” to learn Python. Any introductory course will suffice, as they teach the same things.
The main things you want to learn are:
- Native data structures (dictionaries, lists, sets, and tuples)
- For and while loops
- If-else conditional statements
- Functions and classes
You also want to learn specific scientific computing libraries such as:
- NumPy — Numerical computing and arrays.
- Pandas — Data manipulation and analysis.
- Matplotlib & Plotly — Data visualization.
- scikit-learn — Implementing classical ML algorithms.
Resources:
Timeline: Again, depending on your background, this should take a couple of months. If you know Python already, it will be a lot quicker.
Data structures and algorithms
This one may seem slightly out of place, but if you want to be a machine learning or AI engineer, you must know data structures and algorithms.
This is not only for interviews; it is also used in AI/ML algorithms. You will come across things like backtracking, depth-first search, and binary trees more than you think.
The things to learn are:
- Arrays & Linked Lists
- Trees & Graphs
- HashMaps, Queues & Stacks
- Sorting & Searching Algorithms
- Dynamic Programming
Resources:
- Neetcode.io — Great introductory, intermediate and advanced data structure and algorithm courses.
- Leetcode & Hackerrank — Platforms to practise.
Timeline: Around a month to nail the basics.
Machine learning
This is where the fun begins!
The previous four steps involved getting your foundation ready to tackle machine learning.
In general, machine learning falls into two categories:
- Supervised learning — where we have target labels to train the model.
- Unsupervised learning — when there are no target labels.
The diagram below illustrates this split and some algorithms in each category.
The key algorithms and concepts you should learn are:
- Linear, logistic and polynomial regression.
- Decision trees, random forests and gradient-boosted trees.
- Support vector machines.
- K-means and K-nearest neighbour clustering.
- Feature engineering.
- Evaluation metrics.
- Regularisation, bias vs variance tradeoff and cross-validation.
Resources:
Timeline: This section is quite dense, so it will likely take roughly ~3 months to know most of this information. In reality, it will take years to truly master everything in those resources.
AI and deep learning
There has been a lot of hype around AI since ChatGPT was released in 2022.
However, AI itself has been around as a concept for a long time, dating back in its current form to the 1950s, when the neural network originated.
The AI we refer to at the moment is specifically called generative AI (GenAI), which is actually quite a small subset of the whole AI eco-system as shown below.
As its name suggests, GenAI is an algorithm that generates text, images, audio, and even code.
Until recently, the AI landscape was dominated by two main models:
However, in 2017, a paper called “Attention Is All You Need” was published, introducing the transformer architecture and model, which has since superseded CNNs and RNNs.
Today, transformers are the backbone of large language models (LLMs) and unequivocally rule the AI landscape.
With all this in mind, the things you should know are:
- Neural Networks — The algorithm that really puts AI/ML on the map.
- Convolutional and Recurrent Neural Networks — Still used today quite a bit for their specific tasks.
- Transformers — The current state of the art.
- RAG, Vector Databases, LLM Fine Tuning — These technologies and concepts are crucial to the current AI infrastructure.
- Reinforcement Learning — The third type of learning used to create AI like AlphaGO.
Resources:
- Deep Learning Specialization by Andrew Ng. — This is the follow-on course from the Machine Learning SpecialiSation and will teach all you need to know about Deep Learning, CNNs, and RNNs.
- Introduction to LLMs by Andrej Karpathy (former senior director of AI at Tesla) — learn more about LLMs and how they are trained.
- Neural Networks: Zero to Hero — Starts relatively slow, building a neural network from scratch. However, in the last video, he gets you building your own Generative Pre-trained Transformers (GPT)!
- Reinforcement Learning Course — Lectures by David Silver, a lead researcher at DeepMind.
Timeline: There is a lot here and it’s call quite hard and cutting edge stuff. So around 3 months is probably what it will take you.
MLOps
A model in a Jupyter Notebook has no value, as I have said many times.
For your AI/ML models to be useful, you must learn how to deploy them to production.
Areas to learn are:
- Cloud technologies like AWS, GCP or Azure.
- Docker and Kubernetes.
- How to write production code.
- Git, CircleCI, Bash/Zsh.
Resources:
- Practical MLOps (affiliate link) — This is probably the only book you need to understand how to deploy your machine-learning model. I use it more as a reference text, but it teaches almost everything you need to know.
- Designing Machine Learning Systems (affiliate link) — Another great book and resource to vary your information source.
Research papers
AI is evolving rapidly, so it’s worth staying up to date with all the latest developments.
Some papers I recommend you read are:
You can find a comprehensive list here.
Conclusion
Breaking into AI/ML may seem overwhelming, but it’s all about taking it one step at a time.
- Learn the basics like Python, maths and data structures and algorithms.
- Get your AI/ML knowledge learning supervised learning, neural networks and transformers.
- Learn how to deploy AI algorithms.
The space is ginormous, so it will probably take you about a year to fully grasp everything in this roadmap, and that’s fine. There are literally bachelor’s degrees dedicated to this space, which take three years,
Just go at your own pace, and eventually, you will get to where you want to be.
Happy learning!
Another thing!
Join my free newsletter, Dishing the Data, where I share weekly tips, insights, and advice from my experience as a practicing data scientist. Plus, as a subscriber, you’ll get my FREE Data Science Resume Template!
Dishing The Data | Egor Howell | Substack
Advice and learnings on data science, tech and entrepreneurship. Click to read Dishing The Data, by Egor Howell, a…newsletter.egorhowell.com
Connect with me
Source link
#Ultimate #AIML #Roadmap #Beginners
