Weekly #CrudeOil chart
The #commodity now has formed a descending triangle
Price now bear 200 DMA
sustaining above 93 , it may rise towards 118 as shown,
below 93, likely to drop towards 58 over the winter (economic slowdown scenario)
#OOTT #Bidenflation #inflation

Thats a gg, Satan

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.

Singing the Blues:
Taking Down an Insider Threat

"I had all of the advantages. I was already inside the network. No one suspected me. But they found my hack, kicked me off the network...

...and physically hunted me down."

Many pentests start from the outside, wanting to see how the perimeter might be breached.

This pentest started from the inside. My client wanted to assume they had already been breached, and, if breached, how far could an attacker go.

Could they stop me once I was inside?

So they snuck me in. Disguised me as a new employee. Gave me a work computer, an ID badge, an account in their system... hell, I even had a cubicle w/my assumed name on it.

The only person who knew who I really was was their CISO. Everyone else thought I was Jeremy in Marketing.

During most of the first morning, I completed onboarding, made introductions, and completed menial tasks.

But I had to act quick. I only had a week onsite. I had to hack their network while not raising suspicion.

So I set about it.

You have to understand... most "Internal Pentests" are straight forward. The hard part is breaching the network, but once you're inside, it's a target rich environment. End of Life computers, default passwords, everyone a Local Administrator...

How to trade on Expiry day with Nifty straddle?
(System trading on Expiry)

Triple straddle is one of the best and accurate methods for expiry day

If nifty trading at 15720 then will sell three straddles above 50 points from ATM, below 50 points and ATM.

Ex.
Sell straddle 15700(ATM), 15650(ATM-50), 15750 (ATM+50)

Stoploss of 80% on every leg and run your trade till 3 PM

Once SL hit in any Leg just keep holding other leg till expiry with SL (Never or less chance to hit both SL on expiry day)

Trading timing at 9:20 the Morning