# Logistic Regression and What Classification Really Is Phase 1 continues. Last episode fit a line for regression; this one turns that same linear score into a classifier. **Classification** predicts a discrete category (usually with a probability), not a continuous quantity. **The machine:** keep the linear score z = w·x + b, then squash it with the **sigmoid** σ(z) = 1/(1+e^−z) into a probability p = P(y=1|x). Invert it and you get the **logit**: log(p/(1−p)) = w·x + b, so logistic regression is *linear in log-odds*. Each weight is an **odds ratio** via e^(w_j), why banks and epidemiology still use it for interpretable reason codes. **Why not squared error:** MSE on a sigmoid is non-convex with vanishing gradients on the worst mistakes. The right loss is **log loss / binary cross-entropy**, which is the **negative log-likelihood of a Bernoulli model**, classification's version of MSE-as-Gaussian-MLE. Its gradient collapses to the clean (p − y)·x, the *same form* as linear regression. **Also covered:** the linear decision boundary (a hyperplane) and why logistic regression can't solve XOR; the from-scratch NumPy rebuild (swap identity→sigmoid, MSE→cross-entropy, keep the same update); L2/L1/elastic-net regularization and the **perfect-separation** blow-up; multiclass via one-vs-rest vs **softmax**; a light look at accuracy's trap on imbalanced data (full metrics next episode); history from Verhulst's 1838 growth curve to Berkson's "logit" (1944) and Cox (1958); and the punchline, **logistic regression is one neuron**. **Worked example:** scikit-learn 1.9.0 `LogisticRegression` on Breast Cancer Wisconsin, with StandardScaler, `predict_proba`, `decision_function`, and coefficient reading. ## In the news (June 28–July 5, 2026) - **[Anthropic Claude Sonnet 5](https://www.anthropic.com/news)**, new default for Free/Pro users; 1M context, intro pricing $2/$10 per M tokens; reported ~63.2% SWE-bench Pro. - **Meituan LongCat-2.0** ([VentureBeat](https://venturebeat.com), SiliconANGLE), 1.6T-param MoE coding model, MIT license, reportedly trained and served on non-Nvidia Chinese chips. - **[Google Gemini image models GA](https://blog.google)**, Gemini 3 Pro Image ("Nano Banana Pro"), strong in-image text, SynthID watermark. - Anthropic-led jailbreak severity framework + safety classifier, reported false positives (a precision/recall story). - **Ai2 MolmoMotion**, open vision-language motion model, weights + dataset on [Hugging Face (allenai)](https://huggingface.co/allenai).