TildAlice – Tech Blog for Developers
Python, AI, Deep Learning, Data Analysis, Computer Vision, Reinforcement Learning โ tips and tutorials
Latest Posts
-
Backtest.py to Vectorbt: 3x Faster Parallel Strategies
Migrate from Backtest.py to Vectorbt for 3x faster parallel backtesting. Optimize multi-strategy portfolios with vectorized operations.
-
GraphSAGE vs GAT: Reddit/PPI Inductive Learning 95% F1
Compare GraphSAGE and GAT for inductive learning on Reddit and PPI datasets. Learn which model achieves 95% F1 and why architecture matters.
-
tmux Scroll vs Terminal: 10K Line Benchmark Results
tmux added 15-22% scroll lag across 4 terminals, but terminal choice caused a 3.8x gap. Alacritty rendered 10K lines in 0.48s; Warp took 1.89s.
-
Digital Twin Hype vs Reality: Why Simple FFT Often Wins
A $50K digital twin missed a bearing fault that a $200 FFT setup caught instantly. Here's when physics-based models actually help vs marketing hype.
-
Speculative Decoding vs MoE: 3.2x Cost Gap on Llama 3
Compare Speculative Decoding vs MoE on Llama 3. Discover why one costs 3.2x more and which inference optimization truly delivers better value.
-
FFT Vibration Analysis on Real Bearing Data: Python Setup
Learn FFT vibration analysis on real bearing data with Python. Step-by-step setup using NumPy, SciPy, and Matplotlib for predictive maintenance.
-
Yahoo Finance vs Google vs Seeking Alpha: Portfolio App Test
Yahoo vs Google vs Seeking Alpha portfolio apps tested on 15 stocks. Real data sync delays, international stock gaps, dividend tracking comparison.
-
ARIMA vs Prophet: KOSPI 30-Day Forecast Test Results
Prophet's MAE beat ARIMA by 1% on KOSPI โ but directional accuracy was 7% worse. Here's why error metrics lie and which model actually works for trading.