Antonio Alcántara

Ph.D. candidate in Mathematical Engineering @ UC3M | Data-Driven Optimization, Machine Learning & Forecasting

[Antonio Alcántara @antonioam31]

@_FPalazon 60 papers, tesis dirigidas y experiencia internacional. Si con ese CV no se consigue una acreditación a titular... Sistema podrido 🫠

[Antonio Alcántara @antonioam31]

Very happy with this year's teacher evaluation survey :)

[Antonio Alcántara @antonioam31]

@XWKTED @CalvinTsay @carlos_r84 Well, it depends on the problem we have. But absolutely, very deep networks could lead to surrogate models that are slow to solve. Thank you for your comment.

[Antonio Alcántara @antonioam31]

Excited to share our new working paper! "A Quantile Neural Network Framework for Two-stage Stochastic Optimization" with @carlos_r84 and @CalvinTsay !

Check it out 👀 https://t.co/RYez5wDg8X

A brief summary in 6 tweets 👇

[Antonio Alcántara @antonioam31]

@XWKTED @CalvinTsay @carlos_r84 given first-stage decisions, it will output the distribution of the value function under the uncertainty faced during training (Section 3.1-3.2). If by complex uncertainty you mean we need a bigger network (so more binary variables)...

[Antonio Alcántara @antonioam31]

This work has been the result of my visit to @CogImperial at Imperial College London. Outstanding researchers and even better people. Thank you all for the warm welcome!

[Antonio Alcántara @antonioam31]

Finally, we study the impact of the number of training samples in our NNs, and the effect of the quantile crossing phenomena.

Read more about the case studies here: https://t.co/RYez5wDg8X

[Antonio Alcántara @antonioam31]

Sentirse un poco masoquista y entrar a leer el grupo de @FPUinvestiga ...

[Antonio Alcántara @antonioam31]

@Iberia os he escrito una consulta por MD. Saludos.