In order to understand Artificial intelligence, machine learning, deep learning, we need to start at the very basics.
🧵🧵🧵🧵 👇
Artificial Inteligence = 🖥️ + 🧠
When can you expect your fridge & mobile phone to revolt against you? (Spoiler Alert: Not in the foreseeable future).
Please Do 're-tweet' the Thread & help me spread the knowledge to more people.
![](https://pbs.twimg.com/media/E_2qe66VcAIr--o.png)
In order to understand Artificial intelligence, machine learning, deep learning, we need to start at the very basics.
Person 1: Lets go watch The Dark Knight. Its our marriage anniversary!
Person 2: I dont like going out at night. Its too scary & depressing.
Person 1: Oh, what i meant was watching the movie "The dark Knight".
Person 1: Thanks honey, you're the best :D.
Person 1: As serious as Harvey Dent. As serious as Gautam Gambhir's surname.
Person 2: Okay, I dont know who they are, and obviously I was being sarcastic.
But its more than that. This language, in which computers talk, every string has a unique precise way to understand the 'sentence' (also known as a computer program). No ambiguity. No Ambiguity.
![](https://pbs.twimg.com/media/E_2qfvGUYAQ7etv.png)
This is the stereotypical 1st program which any person new to Computer Science writes in one of the computer programming languages: Python.
As old as our tryst with computers is our yearning to see them do 'intelligent' things. This was the 1st time that humans have turned God. And we wanted to build our prodigy in our own image. Give it the gift of intelligence.
Artificial Intelligence: Any style of programming which enables computers to mimic humans.
![](https://pbs.twimg.com/media/E_2qgftVQAYDrFh.jpg)
![](https://pbs.twimg.com/media/E_2qg33UYAYxi2R.jpg)
![](https://pbs.twimg.com/media/E_2qhSpVgBAzNnk.jpg)
so, through a limited set of atomic or basic programs we can represent an exponential set of outcomes.
![](https://pbs.twimg.com/media/E_2qhv1VgAEHpIi.jpg)
![](https://pbs.twimg.com/media/E_2qiMqVEAgYHqZ.jpg)
![](https://pbs.twimg.com/media/E_2qiuSVQAQhs0D.jpg)
![](https://pbs.twimg.com/media/E_2qjOsVUAE2J3W.jpg)
![](https://pbs.twimg.com/media/E_2qjsPVQAUrsEL.jpg)
![](https://pbs.twimg.com/media/E_2qj-YVIAIIu9W.jpg)
![](https://pbs.twimg.com/media/E_2qkSRUYAE4BRU.jpg)
Good resource: https://t.co/Yn7pV45yip
Now that we can simulate a neuron, what is the next step? Of course, to simulate an interconnected web of neurons! These are what are known as artificial neural networks.
![](https://pbs.twimg.com/media/E_2qkpwVgAExHwV.jpg)
It turns out, single hidden layer feed forward neural networks can approximate ANY* function to arbitrary degree of precision.
*terms & conditions applied.
https://t.co/WFRooX53Sx
While the theorem proves that such networks exist for each function and degree of precision, it says NOTHING about how to find them.
Let me describe this algorithm in few easy steps. I will try to keep it as non-mathematical as possible since many readers might have a non-mathematical background. Some math is unavoidable thought.
![](https://pbs.twimg.com/media/E_2qlg5VUAALzeO.png)
https://t.co/AfxlveCec6
https://t.co/yTZ9W55ncy
We have known about ANNs for many decades. Why then did we have to wait until 2010s before seeing wide ranging applications?
1. Lot of data. If we're learning from data (this is the core of machine learning), we need lot of data! To capture all the variations in data. https://t.co/9x4qnTAWB4
https://t.co/NZDi4fYpU6
https://t.co/tudLnUAhHh
![](https://pbs.twimg.com/media/E_2qnLiVUAMmLrx.png)
I'm been untruthful when i talked about 2 schools of thought in AI. There are several. What predates modern ML and developed alongside the other schools are traditional ML methods. Everybody loves talking about the winners.
https://t.co/ylraMZQips
We did parts of this in college.
It wouldn't be an overstatement to say, everything.
Google, Facebook, Instagram, YouTube, Google Photos, Twitter everything is powered by Deep Learning. Deep Learning is eating software, quite literally.
https://t.co/SXBoEXQuma
You can provide this program english description of what a program should do, and it codes that program. That day might not be far when standard programming is 'deprecated' by Deep Learning models.
https://t.co/7yz7ydUL0a
https://t.co/AprUFaYKJE
https://t.co/SQOQAPUvBL
So many great places. But in my opinion, best one is courses. There are many of them, all of them great. My Masters guide Ravindran sir has free NPTEL courses.
https://t.co/aIjSK13v0F
https://t.co/LwE94cwzoe
for those that prefer the gentler, simpler, easier version.
This is something which has worried many people. The short answer is, not in the foreseeable future. Computers are still quite dumb. They do what they're told. Remember objective functions? That is how you tell a DL model what to do.
More from Sahil Sharma
#neulandlab 🌋🌋 read on to understand this company and its story of transformation. On twitter, 🌋 has become synonymous with neuland, and this is how i will use it throughout the thread. If you like the thread, plz retweet & help fellow investors learn
Starting with a simple primer on the Pharma industry. They make medicines for all of us. The part of the medicine which gives it its potency against the ailment it cures is called API: Active Pharmaceutical
(here is a reference few people will get, but have to make it since I love it :D):
If Guardians of the globe (https://t.co/vH2C1TDbPy) were a medicine, Omniman (https://t.co/0DGb1MAtXN) is the API.
Back to 🌋. There are pharma companies which sell APIs as well as the finished medicine (FDF: Finished Dosage Formulation). Neuland are pureplay API manufacturers. This provides them some advantage since they don't compete with their clients (the FDF makers).
Medicine is diverse. There are medicine which were first discovered in 1920s (https://t.co/6DXNrhOK46), to those being created now (https://t.co/tcr7dAexmO). Naturally then, some of this involves very complex chemistry to manufacture, others are relatively easier to manufacture.
![](https://pbs.twimg.com/media/E3sDL01VEAQMIau.jpg)
Starting with a simple primer on the Pharma industry. They make medicines for all of us. The part of the medicine which gives it its potency against the ailment it cures is called API: Active Pharmaceutical
(here is a reference few people will get, but have to make it since I love it :D):
If Guardians of the globe (https://t.co/vH2C1TDbPy) were a medicine, Omniman (https://t.co/0DGb1MAtXN) is the API.
![](https://pbs.twimg.com/media/E3sDZ5rVIAQJXbq.jpg)
Back to 🌋. There are pharma companies which sell APIs as well as the finished medicine (FDF: Finished Dosage Formulation). Neuland are pureplay API manufacturers. This provides them some advantage since they don't compete with their clients (the FDF makers).
Medicine is diverse. There are medicine which were first discovered in 1920s (https://t.co/6DXNrhOK46), to those being created now (https://t.co/tcr7dAexmO). Naturally then, some of this involves very complex chemistry to manufacture, others are relatively easier to manufacture.
Given the immense interest in this #investment #process thread about how to get investment ideas, thought of creating a logical follow-up 🧵
Given that I have an idea & some sense for valuation & growth, how do i construct a portfolio?
If you like the thread, plz RT.
First thing worth mentioning is that I am only describing my own process. Others can follow a different process which could work for them.
Opposite of a good idea can also be a good idea
Before understanding the process itself, it is important to understand the problems the process is being designed to solve for. What are the key hurdles or troubles I have faced in designing my PF.
Hurdles
1. At any given point in time (in the past) i have only known deeply about 5% of the listed universe in India, or lower. When I say know deeply I mean read an annual report, an investor presentation, a concall, understood why the business is how it is.
2. I only started my corporate life 4 years ago, so my pf size is small compared to yearly savings. This means that I need to deploy large incremental capital every year at least for next few years. This cant be a buy & track pf, this is a keep buying & tracking for a few yrs pf
Given that I have an idea & some sense for valuation & growth, how do i construct a portfolio?
If you like the thread, plz RT.
A lot of people ask: how do you find companies to invest in? very good question.
— Sahil Sharma (@sahil_vi) July 6, 2021
How did you find RACL, Pix?
Creating this ad-hoc thread to share my process.
if you like it, please RT to benefit maximum investors. \U0001f64f
First thing worth mentioning is that I am only describing my own process. Others can follow a different process which could work for them.
Opposite of a good idea can also be a good idea
Before understanding the process itself, it is important to understand the problems the process is being designed to solve for. What are the key hurdles or troubles I have faced in designing my PF.
Hurdles
1. At any given point in time (in the past) i have only known deeply about 5% of the listed universe in India, or lower. When I say know deeply I mean read an annual report, an investor presentation, a concall, understood why the business is how it is.
2. I only started my corporate life 4 years ago, so my pf size is small compared to yearly savings. This means that I need to deploy large incremental capital every year at least for next few years. This cant be a buy & track pf, this is a keep buying & tracking for a few yrs pf
More from Genericlearnings
A list of all the articles that we have created - From Banking to Fiber!
The content in the articles has been written in a very simple language which will help you learn everything about the sector and/or the company!
Do Retweet and help your fellow tweeples learn!
Let's go 👇
Starting with Financials -
A to Z of Banking, all basics of banking explained -
https://t.co/tMfB73CHYs
Top 5 Banks and their strategies -
https://t.co/aivfUtuw9g
Large Bank - HDFC Bank - How did HDFC Bank become HDFC Bank -
Mid Sized Bank - Kotak Mahindra Bank
How did they avoid all NPAs from 1999? What makes Uday Kotak's Concalls a goldmine of information on the Banking sector? Everything explained!
We have given details from 1999! The most comprehensive article ever!
Large NBFC - Consumer Durables Play - Bajaj Finance
From its origins to how it gives 0% EMI to how it earns money from manufacturers - everything explained!
More - Origins, Products, Loan Book, Cross-Selling, Risk management, Concalls of 8 Years,
Gold NBFC - Manappuram Finance
Origins (with fun facts)
Products, 10Y Financials, Business model and how do they make money, How does a gold loan work, Operational efficiency, peer comparison, mgmt commentary, why we don't like the stock, and much
The content in the articles has been written in a very simple language which will help you learn everything about the sector and/or the company!
Do Retweet and help your fellow tweeples learn!
Let's go 👇
Starting with Financials -
A to Z of Banking, all basics of banking explained -
https://t.co/tMfB73CHYs
Top 5 Banks and their strategies -
https://t.co/aivfUtuw9g
Large Bank - HDFC Bank - How did HDFC Bank become HDFC Bank -
Mid Sized Bank - Kotak Mahindra Bank
How did they avoid all NPAs from 1999? What makes Uday Kotak's Concalls a goldmine of information on the Banking sector? Everything explained!
We have given details from 1999! The most comprehensive article ever!
Large NBFC - Consumer Durables Play - Bajaj Finance
From its origins to how it gives 0% EMI to how it earns money from manufacturers - everything explained!
More - Origins, Products, Loan Book, Cross-Selling, Risk management, Concalls of 8 Years,
Gold NBFC - Manappuram Finance
Origins (with fun facts)
Products, 10Y Financials, Business model and how do they make money, How does a gold loan work, Operational efficiency, peer comparison, mgmt commentary, why we don't like the stock, and much
Personal Finance 101 – My learning’s about investing
This topic is for everyone, whether you manage your money yourself or through your advisor, it will go a long way in managing your finances.
Do re-tweet & help us educate retail investors (1/n)
Subscribe to our YouTube for some interesting educational content around Personal Finance - https://t.co/jvgNEDWiAZ
And you can also join our Telegram channel for regular updates – https://t.co/Ekz6I8pDGt (2/n)
(1) Lets start with Life Insurance
Term Insurance is the best way to take an insurance cover & probably the only product to buy in life insurance. Make sure u disclose all the necessary information before taking the insurance. Smoking, Alcohol, any pre-existing deceases etc(3/n)
Have atleast 10-15 times of your annual income as insurance cover
But there are variants of term insurance that you should avoid (4/n)
(A) Term plan with return of premium
For a non-smoker born on the 1st Jan 1985 & policy term 39 years (till age 75), the regular premium for a 1-cr term insurance is 22,157 (inclusive of GST) but with returns of premium is 42670 (inclusive of GST). An increase of 20,513 (5/n)
This topic is for everyone, whether you manage your money yourself or through your advisor, it will go a long way in managing your finances.
Do re-tweet & help us educate retail investors (1/n)
![](https://pbs.twimg.com/media/EwFDLwTXIAg3Oxg.jpg)
Subscribe to our YouTube for some interesting educational content around Personal Finance - https://t.co/jvgNEDWiAZ
And you can also join our Telegram channel for regular updates – https://t.co/Ekz6I8pDGt (2/n)
(1) Lets start with Life Insurance
Term Insurance is the best way to take an insurance cover & probably the only product to buy in life insurance. Make sure u disclose all the necessary information before taking the insurance. Smoking, Alcohol, any pre-existing deceases etc(3/n)
Have atleast 10-15 times of your annual income as insurance cover
But there are variants of term insurance that you should avoid (4/n)
(A) Term plan with return of premium
For a non-smoker born on the 1st Jan 1985 & policy term 39 years (till age 75), the regular premium for a 1-cr term insurance is 22,157 (inclusive of GST) but with returns of premium is 42670 (inclusive of GST). An increase of 20,513 (5/n)