{"id":572,"date":"2025-02-16T16:48:38","date_gmt":"2025-02-16T15:48:38","guid":{"rendered":"https:\/\/noiseonthenet.space\/noise\/?p=572"},"modified":"2025-02-16T16:48:40","modified_gmt":"2025-02-16T15:48:40","slug":"hold-the-line","status":"publish","type":"post","link":"https:\/\/noiseonthenet.space\/noise\/2025\/02\/hold-the-line\/","title":{"rendered":"Hold the Line"},"content":{"rendered":"

\"sam-goodgame-Pe5BC-EDtB4-unsplash.jpg\" Photo by Sam Goodgame<\/a> on Unsplash<\/a>
<\/p>\n\n

We made quite a journey so far! Starting from Jupyter<\/a> and Pandas<\/a> we explored our datasets<\/a> and created independent scripts<\/a>.
<\/p>\n\n

It is now the time to learn the basics of a very powerful tool: Linear Regression.
<\/p>\n\n

Linearity is a key concept in mathematics: it goes very far from the naive idea of a straight line into more abstract concepts like the independence of two effecs into a dynamic system.
<\/p>\n\n

The jupyter notebooks for this series of posts, the datasets, their source and their attribution are available in this Github Repo<\/a>
<\/p>\n\n

<\/a> <\/p>\n

\n

Correlations and linear models<\/h2>\n
\n

A linear model estimate a response from the linear combination of one or more inputs <\/p>\n\n

\"y <\/p> \n\n

\"y <\/p> \n\n

<\/a> given an estimation \"\hat{y} <\/p>\n\n

we look for the best \"\vec{\alpha}\" which minimize all residuals <\/p>\n\n

\"\epsilon(\hat{y}) <\/p> \n\n

\n
import<\/span> pandas as<\/span> pd\n<\/pre>\n<\/div>\n\n
 <\/a> This dataset collects the yearly water and energy consuption estimation per capita in Milan collected by the italian government <\/p>\n\n
 Data is grouped by <\/p>\n\n