BC MACHINE LEARNING
Saved by @zmbnski

Important paper from Google on large batch optimization. They do impressively careful experiments measuring # iterations needed to achieve target validation error at various batch sizes. The main "surprise" is the lack of surprises. [thread]

https://t.co/7QIx5CFdfJ

 The paper is a good example of lots of elements of good experimental design. They validate their metric by showing lots of variants give consistent results. They tune hyperparamters separately for each condition, check that optimum isn't at the endpoints, and measure sensitivity.
They have separate experiments where the hold fixed # iterations and # epochs, which (as they explain) measure very different things. They avoid confounds, such as batch norm's artificial dependence between batch size and regularization strength.
When the experiments are done carefully enough, the results are remarkably consistent between different datasets and architectures. Qualitatively, MNIST behaves just like ImageNet.
Importantly, they don't find any evidence for a "sharp/flat optima" effect whereby better optimization leads to worse final results. They have a good discussion of experimental artifacts/confounds in past papers where such effects were reported.
The time-to-target-validation is explained purely by optimization considerations. There's a regime where variance dominates, and you get linear speedups w/ batch size. Then there's a regime where curvature dominates and larger batches don't help. As theory would predict.
Incidentally, this paper must have been absurdly expensive, even by Google's standards. Doing careful empirical work on optimizers requires many, many runs of the algorithm. (I think surprising phenomena on ImageNet are often due to the difficulty of running proper experiments.)

More from Machine learning

Pratham Prasoon
Pratham Prasoon
@PrasoonPratham
Are you planning to learning machine learning this year?

These are the Python frameworks you need to learn as an absolute beginner.

🧵👇

It can be overwhelming seeing all the machine learning frameworks at once and not even knowing what they do.

I made this thread so that your machine learning journey becomes just a little bit easier.

(2 / 11)

This thread will only focus on Python frameworks.

Why?
Because Python is the most common path into machine learning.

( If you have experience with any other language then learning Python will not be difficult at all. )

(3 / 11)

Here's the entire list:

- Pandas
- Numpy
- Matplotlib
- TensorFlow
- PyTorch
- Keras
- Scikit Learn

Let's see what they do.

(4 / 11)

You'll have to interact with csv files or databases on a regular basis to access data in machine learning, pandas helps you do just that.

The cool thing about Pandas is that it actually doesn't change the original data...

(5 / 11
MACHINE LEARNING
Pratham Prasoon
Pratham Prasoon
@PrasoonPratham
There are several skills needed for learning machine learning that no one talks about.

Here are some of them.
(from what I've learned over the past two years)
🧵👇

This thread aims to introduce to you some of the skills often ignored for learning machine learning but are actually very important.

These skills will enable you to learn concepts quickly and more efficiently in this field.

(2 / 9)

1⃣ Reading

Probably one of the most underrated skills on this list.

In machine learning, you HAVE to read a lot of articles, papers, documentation, and whatnot.

It is mostly due to the theoretical nature of this field that reading is an important skill to have.

(3 / 9)

2⃣ Strong fundamentals in programming

Machine learning is just a lot of programming mixed with math and data.

Having clear programming fundamentals is crucial for this field.

I highly suggest you learn about these if you are using Python for machine learning.👇

(4 / 9)

- Object oriented programming in Python :Classes, Objects, Methods
- Lists & List functions
- Dunder Methods
- List comprehension
- List slicing
- String formatting
- List, Dictionaries & Tuples
- *args,**kwargs

(5 / 9)
MACHINE LEARNING
Advertisement
Alejandro Piad Morffis
Alejandro Piad Morffis
@AlejandroPiad
This is a Twitter series on #FoundationsOfML.

❓ Today, I want to start discussing the different types of Machine Learning flavors we can find.

This is a very high-level overview. In later threads, we'll dive deeper into each paradigm... 👇🧵

Last time we talked about how Machine Learning works.

Basically, it's about having some source of experience E for solving a given task T, that allows us to find a program P which is (hopefully) optimal w.r.t. some metric


According to the nature of that experience, we can define different formulations, or flavors, of the learning process.

A useful distinction is whether we have an explicit goal or desired output, which gives rise to the definitions of 1️⃣ Supervised and 2️⃣ Unsupervised Learning 👇

1️⃣ Supervised Learning

In this formulation, the experience E is a collection of input/output pairs, and the task T is defined as a function that produces the right output for any given input.

👉 The underlying assumption is that there is some correlation (or, in general, a computable relation) between the structure of an input and its corresponding output and that it is possible to infer that function or mapping from a sufficiently large number of examples.
MACHINE LEARNING
Cory Doctorow #BLM
Cory Doctorow #BLM
@doctorow
We hear a lot about Facebook as a platform for manipulation - using machine learning to bypass our critical faculties and trick us into believing things that are bad for us - but the real show is in Facebook's ability to target, not manipulate.

1/


People who hold disfavored views struggle to find one another and mobilize. To find other people that feel the same way as you and make common cause with them to effect political change, you have to reveal your views and suffer social sanction.

2/

Search allows people who hold these views to find one another. If you have a deep feeling about your gender being nonbinary but don't know the words for it, you can search for communities of people who have those words, join them, and discover who you've been all along.

3/

This is why, in this moment, so many ideas are migrating from the fringe to the center: ideas about racial justice, gender identity, alternatives to market systems, etc. People have harbored these views all along, but have held back on expressing them.

4/

Being able to express yourself in private, among people who share your views, is a prelude to going public and putting your case to the wider world in hopes of effecting change.

5/
MACHINE LEARNING
IM POCUS Focus
IM POCUS Focus
@IMPOCUSFocus
A tweetorial on optimal machine settings for B line image acquisition, with a focus on: focal position at pleura, harmonics off, and increased far gain.

Inspired/based on the recent article - https://t.co/vv2xBSZm6N
@cameron_baston @IsaacMatthiasM1

#POCUS #IMPOCUS

1/18

Let's start with some polls
We will stick to low frequency probes here (i.e. curvilinear and phased array - linear has its own purpose for detailed pleural assessment)

Which preset do you prefer for B line assessment?

2/

What depth do you typically set?

3/

After selecting preset and selecting depth, do you ever adjust any other machine settings to optimize B line assessment?

4/

Importantly, there is no right answer to the above questions!

We won't discuss much about preset or probe type. But we will discuss machine settings to optimize B line quality, based on
1) Theory based on physical basis of B lines
2) In vitro study
3) Study in patients

5/
MACHINE LEARNING

You May Also Like

Mo Rajabifard
Mo Rajabifard
@morajabi
👨‍💻 Last resume I sent to a startup one year ago, sharing with you to get ideas:

- Forget what you don't have, make your strength bold
- Pick one work experience and explain what you did in detail w/ bullet points
- Write it towards the role you apply
- Give social proof

/thread


"But I got no work experience..."

Make a open source lib, make a small side project for yourself, do freelance work, ask friends to work with them, no friends? Find friends on Github, and Twitter.

Bonus points:

- Show you care about the company: I used the company's brand font and gradient for in the resume for my name and "Thank You" note.
- Don't list 15 things and libraries you worked with, pick the most related ones to the role you're applying.
-🙅‍♂️"copy cover letter"

"I got no firends, no work"

One practical way is to reach out to conferences and offer to make their website for free. But make sure to do it good. You'll get:

- a project for portfolio
- new friends
- work experience
- learnt new stuff
- new thing for Twitter bio

If you don't even have the skills yet, why not try your chance for @LambdaSchool? No? @freeCodeCamp. Still not? Pick something from here and learn https://t.co/7NPS1zbLTi
You'll feel very overwhelmed, no escape, just acknowledge it and keep pushing.
LIFE
Alex Oluwatobi
Alex Oluwatobi
@alexlobaloba
According to psychologists, there are four types of intelligence:

- Intelligence Quotient (IQ)
- Emotional Quotient (EQ)
- Social Quotient (SQ)
- Adversity Quotient (AQ)

1. IQ is the measure of one's ability to comprehend, solve maths; memorize things & recall subject matters.

2. EQ is the measure of your ability to maintain peace with others; keep to time; be responsible; be honest; respect boundaries; be humble, genuine, and considerate.

3. SQ is the measure of one's ability to build a network of friends & maintain it over a long period of time.

4. AQ is the measure of your ability to go through a rough patch in life and come out without losing your mind. It is a new paradigm and can be used to determine who will give up in face of troubles and may abandon their families.

Now to the very interesting & analytical part👇

People that have higher EQ & SQ tend to go further in life than those with high IQ but low EQ &SQ.

Most schools capitalize on improving IQ level while EQ & SQ are played down.

A man of high IQ can end up being employed by a man of high EQ & SQ even though he has an average IQ.

Your EQ represents your character; your SQ represents your charisma. Give in to habits that will improve these three Qs but more especially your EQ and SQ.

EQ and SQ make one manage better than the other.

Try not only to have a higher IQ but also to have higher EQ and SQ.
ALL
Advertisement
Kanekoa
Kanekoa
@KanekoaTheGreat
*The Evidence Collection*

There is overwhelming evidence of election fraud. Therefore it is the duty of the Executive, Legislative, & Judicial branches to throw out the fraudulent votes & to honor their oath to protect & defend the Constitution of the USA.🇺🇸🦅

A list below.🔻


1/ Navarro Report: The Immaculate Deception - Six Key Dimensions of Election Irregularities -

2/ Analysis of 3,000 U.S. counties shows Joe Biden received 5.6 % more votes in counties using Dominion Voting Systems (Video) -

3/ Antrim Michigan Dominion Voting Systems Forensics Report -

4/ Evidence Grows: ’20 Election Was Rigged -
POLITICS , ALL
\u2728Anshika Sharma
✨Anshika Sharma
@Anshi_________
Low-risk, High-return stocks are what we look for relentlessly

But many of them we know but ignore
some of them don't grab financial headlines

Here are 10 low-risk, high return stocks (that we ignore):

What I mean by “high-return”?

These stocks have OUTPERFORMED the market:

💹Chart with upwards trend Since IPO
💹Chart with upwards trend Over the last 5 Years or since IPO

These are all long-term winners that will keep on winning

What do We Mean by “low-risk”?

A combination of:

❣️Great management
❣️Broad moat
❣️Financially strong
❣️High ROCE
❣️Low Disruption/Dilution/Volatility
❣️Consistent Organic Growth
❣️Assorted Customers/Suppliers
❣️Low debt

1-#CDSL

A depository is a big business and will grow continuously, Its a sustainable business
cause "Market hai toh CDSL bhi hai or rahega"


2- #Dixon

High Management Efficiency with a high ROE of 19.30%
Healthy long-term growth as Net Sales has grown by an annual rate of more than 30%
High Institutional Holdings
LOW RISK HIGH RETURN STOCKS , LOW RISK HIGH RETURN STOCKS FOR LONG TERM , LOW RISK STOCK IDEAS
Seth Abramson
Seth Abramson
@SethAbramson
(THREAD) The Whitaker scandal is deepening, as it becomes clear that Sessions' Russia recusal was a coordinated sham and his firing an act of obstruction—and that his replacement is a White House plant inside the Mueller probe. I hope you'll retweet and read on for more details.


1/ From 2014-17 Whitaker worked for World Patent Marketing—which during his tenure defrauded consumers out of $26 million and was successfully prosecuted by the feds. His involvement in the scam confirmed he had loose morals and that the feds would never want to employ him again.

2/ Despite the seeming impossibility of a man with Whitaker's background getting a job at Justice ever again—he'd been found to have used his former title as a US attorney to fraudulently threaten consumers with valid complaints with criminal penalties—Whitaker found an opening.

3/ Within 60 days of parachuting out of World Patent Marketing as it was being fined $26 million by the feds, Whitaker was working for CNN and telling a fellow attorney panelist that his purpose in working for CNN was to get noticed by one man—Donald Trump—and thereby get a job.

4/ Whitaker spent his time at CNN making ludicrous statements about the Mueller probe: there was no obstruction or collusion, he said; Mueller had no authority to look into any aspect of Trump's finances or to subpoena him, he said. All the while, he hoped Trump was watching him.
POLITICS