{%extends "base.html" %} {% block content %}
Our trend forecasting module uses a hybrid AI approach to predict which fashion aesthetics are likely to rise in popularity. By clustering over 60,000 runway looks into visual categories using BIRCH and PCA, we track how trends emerge, grow, and fade over time. Then, through a combination of LSTM and Transformer models, we forecast the future trajectory of each trend based on historical frequency and momentum. This allows us to model not just what is trending now, but what is poised to make a return—or a debut—on runways in seasons to come.