The **Mathematical Foundations of Data Science **course empowers you to learn machine learning and deep learning (Artificial Intelligence – AI) based on your high school maths concepts.

We help you **to switch your career towards AI** using your existing maths knowledge.

Machine learning and Deep Learning are complex concepts, but you can learn them by **‘joining the dots’** back to your high school maths.

What do we mean by learning AI based on existing maths knowledge?

Essentially, till the year 17/ 18, you learn four components (at least partially) i.e.

- Probability Theory
- Statistics
- Linear Algebra and
- Optimization

These concepts are undoubtedly familiar to you if you have studied a maths/ science-based bachelors or engineering program. These four ideas also underpin machine learning and deep learning. Our work aims to **bridge the gap between these high school maths concepts and data science **(which is a collective term for machine learning and deep learning).

The course is **created by Ajit Jaokar** – who teaches AI at the University of Oxford and LSE. The course is not affiliated to any academic institution.

## How did this approach originate?

I first developed these ideas when I taught my AI courses at the University of Oxford. As a teacher, I noticed that industry professionals often find it easier if I could co-relate complex ideas to simpler concepts. Especially if they already knew these concepts (even if they had learned them 10 or 20 years ago!). Hence, I spend a lot of time ‘connecting the dots.’ Note that this course is not related to any academic institute I am associated with.

**Why is this approach unique?**

Instead of AI, let’s talk instead of Elon Musk!

Elon Musk talks of **first principles thinking**.

That means you boil down a process to its fundamental parts and work backward through each building block to connect to the essential elements. This approach needs more in-depth thinking but provides for deeper learning.

In other words, first principles thinking is the act of identifying the fundamental parts that you know are true and building up from there. When applied in this context (learning AI), we can connect complex ideas to simple ones, especially to ideas you already know.

You may have tried to learn Machine learning and Deep Learning previously with limited success.

In my experience of teaching machine learning and deep learning, three problems inhibit effective learning. Often, these are missed by industry practitioners but are all too familiar to teachers – especially if they teach in class (face to face)

**Cognitive dependencies**– ideas depend on other ideas that need to be explained first.**Cognitive overload**– there is too much information. Most of it is good. But you don’t know which sequence to approach the topic.**First-principles thinking**– i.e., how things tie back to each other.

If these ideas resonate with you, you have come to the right place!

## Details of Topics and Modules

We have two categories of modules: **AF modules (AI fundamentals modules) and MF modules (Maths foundations modules). **In the AI Foundations modules, we first understand machine learning and deep learning in detail, assuming you have had no previous exposure to machine learning or deep learning. We then explore the ten maths foundations modules as below. In each case, we start with a concept you are likely to be already familiar with at the high school level (ex: the equation of a straight line) and then connect it to more complex machine learning or deep learning ideas.

**Mathematics of machine learning**

We start off with Linear algebra and the equation of a straight line. We relate it to perceptrons and multi-layer perceptrons.

**Functions and mathematical modeling**

In this module, we explore machine learning and deep learning as functions relating them to maths

**Probability and inference**

An exploration of probability theory and how it relates to inference

**The maths of deep learning**

An exploration of gradient descent and backpropagation from a maths perspective

**Maths techniques used in feature engineering**

Many maths techniques are used in feature engineering. In this section, we explore feature engineering and co-relate these ideas to maths.

**Model evaluation**

An analysis of model evaluation

**A taxonomy of algorithms**

We look at a technique to classify algorithms from a maths standpoint by considering the sample space or instance space

**Unsupervised learning**

Unsupervised techniques including generative methods

**Bayesian approaches**

In this module, we explore Bayesian approaches from a maths standpoint.

**A recap of ideas**

In this concluding module, we recap all the ideas we have learned.

**Why now and what you get**

- The total price is 99 USD + taxes as applicable
- If you sign up now, as a limited promotional offer – we offer
**monthly zoom mentoring**sessions for the group (once a month). - Sign up now for only 99 USD.
- When you sign up, you get the AI foundations modules – the next third in the next month and final third month after (spanning Sep Oct Nov Dec)
- You get lifetime access to content and videos.
- You can post questions online, which I can answer – however, considering the accessible price point – responses will not be instant. But I will respond to all questions related to the course within the week.
- The material is based on my prior teaching offline (but we are not affiliated to any educational institution)
- I intend to keep this group small and selective so that we can learn from it.
- This pricing is a fixed cost, and it’s intentionally low to keep the material accessible globally. There are no extras.

**Frequently Asked Questions**

Yes – it is. I have tried to keep it affordable globally. But I am also trying to keep the group size small. Hence, the numbers will be limited. The price is all-inclusive, and there are no other extras.

There is no limit to the access. The material would be available to you at all times after even after the course completion.

Since you have access to the content, you can take longer than three months if you like.

No. The course is not associated with any academic institution.

There is nothing to read before the course, and the course assumes that you do not have existing knowledge of machine learning or deep learning.

Yes, 😊 Maths – when studied in this way – applies to specific problems. Even if you never liked it before, you may be surprised to see how applied maths can help you understand machine learning, deep learning, and AI.

I am happy to discuss, but probably this approach may not be ideal for you if you have no familiarity with topics like Linear Algebra at some point in your life.

No, the course does not involve coding. There are many courses taking a coding-based approach. We focus only on a maths-based approach where you understand AI from maths foundations.

No, where needed – we can refer to them on the Web if you need to brush up.

No. It is for anyone who has studied some maths at high school level at any point in their life.

Possibly. But if you are under 18, some special considerations may apply in terms of engagement. Please contact us.

Yes. You can easily pick it up.

I have never taught this course in this format(online). So, this is genuinely the first time I am running this course online. I could ask my students who I taught face to face – but the format is different – and it would not apply. By keeping the first batch small, I am hoping that you could be the first reference customer!

We offer the course free to anyone who is on the autistic/ Asperger spectrum. Please contact me for this from the contact form.

Based in London, Ajit’s work spans research, entrepreneurship and academia relating to artificial intelligence (AI) with Edge Computing, 5G and Cybersecurity.

He is the course director of the course: Artificial Intelligence: Cloud and Edge Implementations. Besides this, he also conducts the University of Oxford courses: AI for Cybersecurity and Computer Vision.

Ajit works as a Data Scientist through his company feynlabs – focusing on building innovative early stage AI prototypes for domains such as cybersecurity, robotics and healthcare.

Besides the University of Oxford, Ajit has also conducted AI courses in the London School of Economics (LSE), Universidad Politécnica de Madrid (UPM) and as part of the The Future Society at the Harvard Kennedy School of Government.

He is also currently working on a book to teach AI using mathematical foundations at high school level.

His PhD research is based on AI and Affective Computing (how AI interprets emotion).