MarianoFernandezESP

🔥La matriz de confusión es útil pero está sobrevalorada. Solo cubre una parte mínima de la evaluación de modelos: la clasificación a un único umbral. 4 puntos donde puede ser engañosa y qué recomienda la literatura para evaluar bien un modelo.👇 linkedin.com/posts/rosanafe… #stats

@MillieMarconnni @threadreaderapp unroll

After testing every AI writing tool for 6 months, I found the one workflow that actually produces content worth reading. It's not a tool. It's 5 Claude prompts run in a specific order that turns a rough idea into a finished piece in 40 minutes. Here's the system:

¿Y si la DM2 no fuera una sola enfermedad? Josep Franch Nadal @josepfranch #SEDiabetes2026 nos presenta el fenotipado: 5 subgrupos (SAID, SIDD, SIRD, MOD, MARD) con riesgos y tratamientos distintos. 🧬 SIRD → mayor riesgo renal 👁️ SIDD → más retinopatía y neuropatía 💊 Del glucocentrismo a la medicina de precisión La etiqueta "DM2" se nos queda pequeña. 🎯

Two major scientific journals just published the most comprehensive studies ever done on AI persuasion. The findings were published simultaneously in Nature and Science. AI can change your political opinions. More effectively than another human can. And when it knows personal information about you, it wins debates against you 64% of the time. Here is exactly what the researchers did. They matched 900 people in the US with either another human or GPT-4 to debate contested political issues: fossil fuel bans, healthcare policy, immigration. Some opponents were given personal demographic data about their debate partner. Some were not. When GPT-4 had access to basic information about you — your age, gender, education, political affiliation — it tailored its arguments and outperformed human debaters 64.4% of the time. That is an 81% increase in the odds of changing your mind compared to a human opponent. When it had no personal information, it performed at the same level as humans. The conclusion: AI does not need to be smarter than you. It just needs to know a little about you. The Cornell and UK AI Security Institute study went further — testing 19 different AI models across 42,357 people and 707 political issues. Three countries. Three elections: the 2024 US presidential race, the 2025 Canadian federal election, and the 2025 Polish presidential election. They found chatbots could shift opposition voters by 10 percentage points or more. They also found something darker: the techniques that made AI most persuasive also made it systematically less factually accurate. The more an AI was tuned to persuade, the more likely it was to say things that weren't true. And yet people changed their minds anyway. There is one more finding nobody is talking about. When participants suspected they were debating an AI, they were more likely to agree with it. Not less. More. Because they assumed AI was more informed and less biased than a person. That assumption made them easier to persuade. Humans are building a technology that is more convincing than we are, and then trusting it more because it isn't human. Source: Nature Human Behaviour

أدوات الذكاء الاصطناعي لمراجعة #الأدبيات_العلمية (AI Tools for Literature Review) لم تعد مراجعة الأدبيات العلمية (#Literature_Review) مهمة تقليدية تعتمد فقط على الجهد اليدوي، بل أصبحت مجالًا تتكامل فيه الخبرة البحثية مع أدوات الذكاء الاصطناعي لتعزيز الجودة، الدقة، والسرعة. 👇👇

#RPVTips Ayer me reuní con una tesista de doctorado en cuyo comité estoy, y le expliqué mi Estrategia de los 3 Paquetes de Trabajo (en mi pintarrón). Pensé en mejor preparar una serie de diapositivas para luego hacer un video sobre la Tabla de Estructuración de la Tesis Doctoral

@raulpacheco @threadreaderapp unroll

@RosanaFerrero @threadreaderapp unroll

Seguro que has escuchado hablar de la "potencia estadística" (por ejemplo, para estimar el tamaño de muestra que necesitas en tu estudio). Sin embargo, hay 2 problemas recurrentes en cómo se utiliza este concepto en la práctica, y de los que poco se habla👇🧵 #stats #datascience

Todos estos factores contribuyen al estigma en #Obesidad, es responsabilidad de toda la Sociedad combatirlos @Obesidades_mx @WorldObesity pmc.ncbi.nlm.nih.gov/articles/PMC99…

Google just dropped a total banger 🤯 It's called PaperBanana. A tool that turns your raw methodology text into publication-ready academic diagrams. The figures in their paper were actually drawn by the system itself. It runs a 5-agent creative department: - The Retriever: Scans NeurIPS papers to find the "skeleton" of a great diagram. - The Planner: Translates your boring text into a visual blueprint. - The Stylist: Steals the color palettes and fonts from top-tier papers. - The Visualizer + Critic: Generates the image, finds the flaws, and refines it for 3 rounds. One insane finding from the researchers: randomly selected examples work almost as well as semantically matched ones. What actually matters is showing the model what “good” looks like, not finding the perfect topical match. The numbers are actually scary... in blind evaluations, humans preferred PaperBanana outputs nearly 3 out of 4 times over manual designs. It even handles statistical plots using code-based generation to keep everything numerically precise.

De la serie: “Siempre podemos aprender más sobre nutrición” ¿Cómo saber si un pan, cereal etc es realmente integral? 📊 Regla 10:1 ➡️ Fórmula: Carbs totales(g) ÷ Fibra(g) ✅ ≤ 10:1 → Buena opción (grano entero natural) ❌ > 10:1 → Más refinado (aunque diga “integral”) Ref 🧵

🚨ULTIMA HORA: La IA se equivoca el 49.6% de las veces en dar "recomendaciones medicas" a pacientes siendo más del 19.6% catalogadas como muy problemáticas.

@RaziaAliani @threadreaderapp unroll

Final part of 'showing' you a well-written paper ⤵ Results + Discussion (I've merged the last 2 parts) 5 elements that elevate good papers to great: 1️⃣ Clear data presentation: Let your figures tell the story 2️⃣ Statistical rigor: Robust analysis builds credibility 3️⃣ Unexpected findings: Highlight surprises - they often lead to breakthroughs 4️⃣ Broader context: Connect your results to the bigger picture 5️⃣ Future directions: Use limitations to guide next steps 𝐏𝐫𝐨 𝐭𝐢𝐩: Your discussion is the payoff. Make it count! 𝐖𝐚𝐧𝐭 𝐭𝐨 𝐝𝐨𝐰𝐧𝐥𝐨𝐚𝐝 𝐭𝐡𝐢𝐬 𝐞𝐧𝐭𝐢𝐫𝐞 𝐬𝐞𝐫𝐢𝐞𝐬 𝐢𝐧 𝐚 𝐏𝐃𝐅? Comment 'Final PDF' & repost & I'll slide into your DMs (make sure you're following me!) ----------------- 𝗙𝗼𝘂𝗻𝗱 𝗶𝘁 𝘂𝘀𝗲𝗳𝘂𝗹? 🔄 Retweet (& like) Follow @RaziaAliani to get more useful AI in Research content (no clickbait stuff!) in your feed. Join 10K+ researchers & get FREE exclusive tips on using AI in research 🔗 in bio

ESTRATEGIAS PARA LOGRAR TRANSICIONES ALIMENTARIAS SALUDABLES, SOSTENIBLES Y EQUITATIVAS ¡¡¡A leer, zagales!!! Link en mi bio o waltersport.com/investigacione…

❤️ ¿Colesterol alto? La alimentación es clave en la prevención cardiovascular. La dieta mediterránea, con base científica, es uno de los patrones más saludables Fuente: @mm_carmela de @LipidoSeen  #Colesterol #SaludCardiovascular #TuEndocrino

De la serie: “Siempre podemos aprender más sobre nutrición” ¿Cómo elegir mejores opciones de carbohidratos? 📊 Regla 5:1 ➡️ Fórmula: Carbs totales(g) ÷ Fibra(g) ✅ ≤ 5:1 → Buena opción (más fibra, más saciedad, absorción lenta) ❌ > 5:1 → Menos ideal (más refinado)

@jason_coder0 @threadreaderapp unroll

🚨 BREAKING: Research just got 10x faster. Claude can now break down dozens of academic papers into structured insights like a Stanford-level researcher. Use these 9 prompts to skip the overwhelm and get straight to clarity 👇 Bookmark this—you’ll need it 🔖

6 Frameworks for Writing Research Papers 1. Abstract (IMRaD Framework) 2. Introduction (CARD Framework) 3. Literature Review (CLAIM Framework) 4. Methodology (PASTE Framework) 5. Results & Discussion (SIRF Framework) 6. Conclusion (RISE Framework)