gradient descent makes a pass over all points to make one gradient step. We can instead approximate gradients on a minibatch of data. This is the idea behind stochastic-gradient-descent.

\begin{equation} \theta^{t+1} = \theta^{t} - \eta \nabla_{\theta} L(f_{\theta}(x), y) \end{equation}

this terminates when theta differences becomes small, or when progress halts: like when \(\theta\) begins going up instead.

we update the weights in SGD by taking a single random sample and moving weights to that direction.

the stock indicies

Stock Issues are policy debate doctrines which divides the debate into 5 subtopical ideas.

Wikipedia

1. Harms: what are the problems in the status quo?

2. Inherency: what are these problems not already being solved? (Or not already being solved in the best way?)

3. Significancy: comparing the advantages and disadvantages of the status quo and your proposed solution, why is the proposed solution more worthy than the status quo?

The Ws:

1. Why this? Why is your proposed solution the best (most effective, or most feasible, or fastest, etc.) one?

• Around 20,000 stocks valued at $47 Trillion
• Only about 2,000 matter
• Transaction frequency is high, liquidity is generally low — grade sizes are small
• Roughly 59 places to trade stock (exchanges + darkpools)