القائمة
Walk-forward stacking ensemble that trains base models on expanding windows and a meta-learner on their out-of-fold predictions across time
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Memory-efficient Keras generator that streams sharded CSV files in chunks, renders strokes to images on-the-fly, and yields batches for training on datasets too large for memory
Stack 1D convolutions for local feature extraction before bidirectional LSTMs to classify variable-length stroke sequences into hundreds of doodle categories
Partition a large dataset into N balanced shards using integer key modulo arithmetic for reproducible, class-interleaved splits across CSV files
Normalize raw stroke coordinates to 0-255 range, resample at uniform arc-length spacing, then apply Ramer-Douglas-Peucker simplification
Render stroke sequences to grayscale images with temporal intensity encoding where earlier strokes are brighter and later strokes fade to encode drawing order
Score each face in a video as anomalous by computing L2 distance from the embedding centroid of all faces, then convert to probability via logistic function
| name | timeseries-expanding-window-stacking |
| description | Walk-forward stacking ensemble that trains base models on expanding windows and a meta-learner on their out-of-fold predictions across time |
Standard K-fold stacking leaks future information in time series. Expanding-window stacking respects temporal order: base models are trained on data up to time t, predict t+1, and the meta-learner trains on these walk-forward out-of-fold predictions. This produces a properly validated stacking ensemble for forecasting tasks.
import numpy as np
from sklearn.linear_model import Ridge
def expanding_stack(df, base_models, time_col="month", target="y"):
periods = sorted(df[time_col].unique())
oof_preds = np.zeros((len(df), len(base_models)))
for i, cutoff in enumerate(periods[12:], start=12): # min 12 months training
train_mask = df[time_col] < cutoff
val_mask = df[time_col] == cutoff
for j, model in enumerate(base_models):
model.fit(df.loc[train_mask, features], df.loc[train_mask, target])
oof_preds[val_mask, j] = model.predict(df.loc[val_mask, features])
# Train meta-learner on OOF predictions
valid_mask = df[time_col] >= periods[12]
meta = Ridge(alpha=1.0)
meta.fit(oof_preds[valid_mask], df.loc[valid_mask, target])
return meta, oof_preds