135135 < h1 > 📊 Mathematics for Data Science & Machine Learning</ h1 >
136136 < p > A comprehensive, beautifully rendered reference for formulas and concepts.</ p >
137137 < div class ="hero-stats " aria-hidden ="true ">
138- < div class ="stat-box "> < span class ="stat-number "> 15 </ span > < span class ="stat-label "> Topics</ span > </ div >
138+ < div class ="stat-box "> < span class ="stat-number "> 17 </ span > < span class ="stat-label "> Topics</ span > </ div >
139139 < div class ="stat-box "> < span class ="stat-number "> 200+</ span > < span class ="stat-label "> Formulas</ span > </ div >
140140 < div class ="stat-box "> < span class ="stat-number "> ∞</ span > < span class ="stat-label "> Applications</ span > </ div >
141141 </ div >
@@ -159,6 +159,8 @@ <h1>📊 Mathematics for Data Science & Machine Learning</h1>
159159 < button class ="nav-btn " onclick ="scrollToSection('svm') "> SVM</ button >
160160 < button class ="nav-btn " onclick ="scrollToSection('decision-trees') "> Decision Trees</ button >
161161 < button class ="nav-btn " onclick ="scrollToSection('bias-variance') "> Bias-Variance</ button >
162+ < button class ="nav-btn " onclick ="scrollToSection('loss-functions') "> Loss Functions</ button >
163+ < button class ="nav-btn " onclick ="scrollToSection('feature-engineering') "> Feature Eng.</ button >
162164 </ div >
163165 < div class ="expand-collapse-all ">
164166 < button class ="expand-collapse-btn " onclick ="expandAllSections() "> 📂 Expand All</ button >
@@ -916,6 +918,83 @@ <h2 id="bv-title" class="section-title">Bias-Variance Tradeoff</h2>
916918 </ div >
917919 </ section >
918920
921+ <!-- 16. QUICK REFERENCE: LOSS FUNCTIONS -->
922+ < section id ="loss-functions " class ="section " aria-labelledby ="loss-title ">
923+ < div class ="section-header " onclick ="toggleSection(this) ">
924+ < div class ="section-number "> 16</ div >
925+ < h2 id ="loss-title " class ="section-title "> Quick Reference: Loss Functions</ h2 >
926+ < div class ="section-toggle "> ▼</ div >
927+ </ div >
928+ < div class ="section-content ">
929+ < div class ="subsection ">
930+ < div class ="subsection-title "> 📉 Regression Losses</ div >
931+ < table class ="formula-table ">
932+ < tr > < th > Loss</ th > < th > Formula</ th > < th > Use Case</ th > </ tr >
933+ < tr > < td > MSE</ td > < td > $$L=\frac{1}{n}\sum_{i=1}^n(y_i-\hat{y}_i)^2$$</ td > < td > Standard regression</ td > </ tr >
934+ < tr > < td > MAE</ td > < td > $$L=\frac{1}{n}\sum_{i=1}^n|y_i-\hat{y}_i|$$</ td > < td > Robust to outliers</ td > </ tr >
935+ < tr > < td > Huber</ td > < td > $$L=\begin{cases}\frac{1}{2}(y-\hat{y})^2&|y-\hat{y}|\leq\delta\\\delta|y-\hat{y}|-\frac{1}{2}\delta^2&\text{otherwise}\end{cases}$$</ td > < td > Combines MSE & MAE</ td > </ tr >
936+ </ table >
937+ </ div >
938+
939+ < div class ="subsection ">
940+ < div class ="subsection-title "> 🎯 Classification Losses</ div >
941+ < table class ="formula-table ">
942+ < tr > < th > Loss</ th > < th > Formula</ th > < th > Use Case</ th > </ tr >
943+ < tr > < td > Binary Cross-Entropy</ td > < td > $$L=-[y\log(\hat{y})+(1-y)\log(1-\hat{y})]$$</ td > < td > Binary classification</ td > </ tr >
944+ < tr > < td > Categorical Cross-Entropy</ td > < td > $$L=-\sum_i y_i\log(\hat{y}_i)$$</ td > < td > Multi-class classification</ td > </ tr >
945+ < tr > < td > Hinge Loss</ td > < td > $$L=\max(0,1-y\cdot\hat{y})$$</ td > < td > SVM classification</ td > </ tr >
946+ </ table >
947+ </ div >
948+ </ div >
949+ </ section >
950+
951+ <!-- 17. FEATURE ENGINEERING -->
952+ < section id ="feature-engineering " class ="section " aria-labelledby ="fe-title ">
953+ < div class ="section-header " onclick ="toggleSection(this) ">
954+ < div class ="section-number "> 17</ div >
955+ < h2 id ="fe-title " class ="section-title "> Feature Engineering</ h2 >
956+ < div class ="section-toggle "> ▼</ div >
957+ </ div >
958+ < div class ="section-content ">
959+ < div class ="subsection ">
960+ < div class ="subsection-title "> 📊 Normalization Techniques</ div >
961+ < table class ="formula-table ">
962+ < tr > < th > Method</ th > < th > Formula</ th > < th > Range</ th > </ tr >
963+ < tr > < td > Min-Max Scaling</ td > < td > $$x'=\frac{x-\min}{\max-\min}$$</ td > < td > [0, 1]</ td > </ tr >
964+ < tr > < td > Z-Score Normalization</ td > < td > $$x'=\frac{x-\mu}{\sigma}$$</ td > < td > ~ [-3, 3]</ td > </ tr >
965+ < tr > < td > Max Abs Scaling</ td > < td > $$x'=\frac{x}{|\max|}$$</ td > < td > [-1, 1]</ td > </ tr >
966+ </ table >
967+ </ div >
968+
969+ < div class ="subsection ">
970+ < div class ="subsection-title "> 🔢 Polynomial Features</ div >
971+ < div class ="formula-card ">
972+ < div class ="formula-name "> Degree 2</ div >
973+ < div class ="formula "> $$[x_1, x_2] \rightarrow [1, x_1, x_2, x_1^2, x_1 x_2, x_2^2]$$</ div >
974+ </ div >
975+ < div class ="formula-card ">
976+ < div class ="formula-name "> Degree 3</ div >
977+ < div class ="formula "> $$[x_1, x_2] \rightarrow [1, x_1, x_2, x_1^2, x_1 x_2, x_2^2, x_1^3, x_1^2 x_2, x_1 x_2^2, x_2^3]$$</ div >
978+ </ div >
979+ < div class ="info-box ">
980+ < strong > 💡 Tip:</ strong > Polynomial features help capture non-linear relationships in data, but be careful of overfitting with high degrees.
981+ </ div >
982+ </ div >
983+
984+ < div class ="subsection ">
985+ < div class ="subsection-title "> 🏷️ Encoding Categorical Variables</ div >
986+ < div class ="formula-card ">
987+ < div class ="formula-name "> One-Hot Encoding</ div >
988+ < div class ="formula "> $$\text{Category } c \rightarrow [0, 0, \ldots, 1, \ldots, 0] \text{ (1 at position } c\text{)}$$</ div >
989+ </ div >
990+ < div class ="formula-card ">
991+ < div class ="formula-name "> Label Encoding</ div >
992+ < div class ="formula "> $$\text{Categories } \{A, B, C\} \rightarrow \{0, 1, 2\}$$</ div >
993+ </ div >
994+ </ div >
995+ </ div >
996+ </ section >
997+
919998 <!-- CONTRIBUTOR SECTION -->
920999 < section id ="contributor " class ="section " style ="text-align:center; ">
9211000 < div class ="section-header " onclick ="toggleSection(this) ">
0 commit comments