Fashion Design Prediction
Yusan discusses the surprising success of the model in predicting fashion designs using LSTM, highlighting the challenge of making the results interpretable. The conversation delves into the potential of deep learning techniques and the need to avoid overfitting in predicting designer styles.In this clip
From this podcast
Data Skeptic
Fashion Predictions
Related Questions
Can you expand on smart recommendation systems in the fashion industry as discussed in the episode Web Scale Engineering for Machine Learning with Sharath Rao - #40 and the clip Model Context Sensitivity
Is that perhaps, you know, the levels you're hitting, as you said, they're better than random. Maybe they're not perfect, but I don't think perfect is the bar. Do you have any sense of where the ceiling is? If this was somebody's or a team of analysts full time job, would they be outperforming the model by a lot?
What can make model predictions in design more interpretable?