aestrad7 Goto Github PK

classification_basic_v1

This exercise arises from the need to explain to a client what conditions a database must meet in order to be modeled in GCP autoML in a classification model. It consists of 3 parts, the general exploration of the variables in shape conditions, the exploration in conditions of linear and / or non-linear correlation and the prediction with a classification model by LogisticRegression from sklearn. Although it is not my favorite classification technique, in this case it meets both the criteria of ease to explain to the client the business philosophy behind the process.

mail-reader

Lector y resumidor de correos

nfs-npl-globant-test-2024

purchase-card-transactions_glbnt_testt

In this dataset you have a collection of purchase card transactions for the Birmingham City Council. This is a historical dataset, you’re able to perform any of the following tasks: (Clustering) Discovering profiles (whether the case) or unusual transactions (anomalies detection) (Forecasting) Try to guess future transactional behaviors. For instance, what would be the next purchase? Expenditures forecasting? (Creativity) State a problem. It’s up to you defining the time window in which your analysis will take place.

readmitted-patients-prediction

Realizar un modelo de predicción que determine si un paciente será readmitido en el hospital en menos de 30 días.

real-estate-basic-pred

reinforcement_poker

La idea es setear todo el dominio de un juego de poker para simular una partida normal. Luego entrenar un algoritmo de aprendizaje no supervisado para crear estrategias de poker.