robintux’s gists

robintux / perceptron_1957_1969.html

Created March 30, 2026 01:44

	<!DOCTYPE html>
	<html lang="es">
	<head>
	<meta charset="UTF-8">
	<meta name="viewport" content="width=device-width, initial-scale=1.0">
	<title>El Perceptrón: El Big Bang de la IA</title>
	<script src="https://cdn.tailwindcss.com"></script>
	<!-- Lucide Icons -->
	<script src="https://unpkg.com/lucide@latest"></script>
	<!-- KaTeX for Math -->

robintux / Descomposición_clásica_del_error_de_generalización.py

Created March 27, 2026 17:01

	"""
	Demostración empírica de la descomposición sesgo-varianza
	Simulación de Monte Carlo para estimar cada componente
	"""

	import numpy as np
	import matplotlib.pyplot as plt
	from sklearn.linear_model import LinearRegression, Ridge
	from sklearn.tree import DecisionTreeRegressor
	from sklearn.ensemble import RandomForestRegressor

robintux / demo_cybenko.py

Last active March 27, 2026 16:24

	# ============================================================================
	# DEMOSTRACIÓN: Aproximación Universal con Keras
	# ============================================================================

	import numpy as np
	import tensorflow as tf
	from tensorflow import keras
	from tensorflow.keras import layers
	from tensorflow.keras.callbacks import EarlyStopping
	import matplotlib.pyplot as plt

robintux / solucion_analitica_poisson_2d.md

Created March 17, 2026 15:43

function poisson_2d_analytical(x, y, f0; n_terms=50)
    phi = zeros(length(x), length(y))
    for (i, xi) in enumerate(x), (j, yj) in enumerate(y)
        sum_val = 0.0
        for m in 1:2:n_terms  # Solo términos impares
            for n in 1:2:n_terms  # Solo términos impares
                coef = (16 * f0) / (π^4 * m * n * (m^2 + n^2))
 sum_val += coef * sin(m * π * xi) * sin(n * π * yj)

robintux / generacion_respuestas.py

Created February 25, 2026 19:46

	def generar_respuesta(prompt, max_length=200, num_beams=4):
	"""Genera respuesta del modelo"""

	inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=512).to(device)

	with torch.no_grad():
	outputs = model.generate(
	**inputs,
	max_length=max_length,
	min_length=20,

robintux / demostacion_empirica.py

Created February 22, 2026 21:42

	import numpy as np
	import matplotlib.pyplot as plt
	from sklearn.pipeline import Pipeline
	from sklearn.preprocessing import PolynomialFeatures
	from sklearn.linear_model import LinearRegression
	from sklearn.metrics import mean_squared_error
	from sklearn.model_selection import train_test_split

	# Configuración de estilo para publicaciones académicas
	plt.style.use('seaborn-v0_8-whitegrid')

robintux / utils.py

Last active February 21, 2026 05:20

	import numpy as np
	import copy
	import time
	import signal

	import matplotlib.pyplot as plt
	from graphviz import Digraph

	from pandas import DataFrame
	from sklearn.metrics import r2_score, mean_absolute_error, mean_squared_error, log_loss

robintux / word_2_vec_simple.py

Created February 20, 2026 19:16

	import numpy as np
	from collections import Counter
	import matplotlib.pyplot as plt
	from sklearn.decomposition import PCA

	class Word2VecSimple:
	"""
	Implementación simplificada de Word2Vec (Skip-gram) para fines educativos.
	"""

robintux / prueba_t_para_una_muestra.py

Created February 17, 2026 18:43

	import numpy as np
	from scipy import stats
	import matplotlib.pyplot as plt
	from typing import Tuple, Callable

	class PowerFunctionOneSampleT:
	"""
	Calcula y visualiza la función de poder para prueba t de una muestra.

	Conexión con Sesión 1:

robintux / temporal_estrat_train_test.py

Created February 15, 2026 22:19

	def temporal_train_test_split(df, predictors, target, test_size=0.2):
	"""
	División temporal que respeta:
	1. Orden cronológico (no leakage)
	2. Integridad de series por país (no partir series arbitrariamente)
	3. Balance de países entre train/test
	"""
	# Ordenar por año
	df_sorted = df.sort_values('Year').reset_index(drop=True)

Abraham Zamudio Chauca robintux