Alexei Ovsyannikov<p>A fascinating data science dilemma</p><p>An ai model notices shark attacks strongly correlate with ice cream sales. The predictive power is great 😸 </p><p>But it completely misses the actual cause: hot weather</p><p>So the classic question: is a strong correlation enough for a model or is understanding true causality essential? What's your take?</p><p><a href="https://pkm.social/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://pkm.social/tags/ml" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ml</span></a> <a href="https://pkm.social/tags/causality" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>causality</span></a> <a href="https://pkm.social/tags/correlation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>correlation</span></a></p>