Monday, March 29, 2021

So far an interesting day. Tried to get some surfboard shaping space at Maketory. Will have to see if this works out, there is a shipping container that could be a good space perhaps for even two people. Will have to see. Hopefully the price is good.

On another note, started working with GWR (Geographically Weighted Regression) to analyze expected AirBnB prices in Berlin. This eventually will be targeted for estimating rent prices across San Diego. However I will say it seems like the expected price as outputted by the regression is quite off.

From what I can tell it is giving almost all weight to the geographically close neighbors, but not taking into account the other important features. I am not quite sure how to correct for this yet. This is something I would like to discuss with Kevin and see if he has any insight that I do not. I am also curious if I can naively do some kind of ensemble technique and see if that helps or not.

I have a feeling I will have to spend a lot more time with the data to get something meaningful out of it. I was hoping the machine learning techniques would reduce the search space dramatically. I think that it should, but I don't have the right insight into the problem or the space yet to really grasp.

I am thinking about taking a break from this problem and instead moving to finish up the Elixir challenges on At the worst I learn more about Elixir and the best case I also win $100 of BTC.

Hmmm okay.

Well Elixir is proving to be more challenging than expected. Doing the Diffe-Hellman key exchange was trivial once I understood the syntax for dealing with Binaries.

Basically a Binary is something in the format <<>>. Every element within the braces will be a uint8. So you might get <<124, 213, 20, 42>>. This would be a Binary containing 4 bytes.

Suppose I wanted to interpret this as two uint16's. I can do the following.

iex(1)> <<i1::integer-unsigned-size(16), i2::integer-unsigned-size(16)>> = <<124, 213, 20, 42>>
<<124, 213, 20, 42>>

iex(2)> i1

iex(3)> i2

Similarly I could interpret this as an int32

iex(4)> <<i1::integer-size(32)>> = <<124, 213, 20, 42>>
<<124, 213, 20, 42>>

iex(5)> i1

Dealing with multiprocessing is much more challenging for me. I am not quite caught up with all the techniques in general. Plus I am still relatively unfamiliar with the syntax which is making it even more challenging.