minor edits

lectures/notes/14_feature-engineering-selection.ipynb

@@ -83,7 +83,7 @@
     "From this lecture, students are expected to be able to:\n",
     "\n",
     "- Explain what feature engineering is and the importance of feature engineering in building machine learning models.  \n",
-    "- Carry out preliminary feature engineering on numeric and text data. \n",
+    "- Carry out preliminary feature engineering on numeric data. \n",
     "- Explain the general concept of feature selection. \n",
     "- Discuss and compare different feature selection methods at a high level. \n",
     "- Use `sklearn`'s implementation of model-based selection and recursive feature elimination (`RFE`)"
@@ -5634,7 +5634,9 @@
     "\n",
     "- Consider a supervised classification task of predicting whether someone has particular genetic variation (SNP)\n",
     "\n",
-    "<img src='../img/sex_mom_dad.png' width=\"600\" height=\"600\" />\n",
+    "![](../img/sex_mom_dad.png)\n",
+    "\n",
+    "<!-- <img src='../img/sex_mom_dad.png' width=\"600\" height=\"600\" /> -->\n",
     "\n",
     "- True model: You almost have the same value as your biological mom."
    ]
@@ -5654,7 +5656,9 @@
     "    - (SNP != biological mom) with very low probability \n",
     "    \n",
     "\n",
-    "<img src='../img/SNP.png' width=\"400\" height=\"400\"/>\n"
+    "![](../img/SNP.png)\n",
+    "\n",
+    "<!-- <img src='../img/SNP.png' width=\"400\" height=\"400\"/> -->\n"
    ]
   },
   {
@@ -5672,7 +5676,9 @@
     "- Dependence, collinearity for linear models\n",
     "    - If a feature can be predicted from the other, don't know which one to pick. \n",
     "\n",
-    "<img src='../img/sex_mom_mom2_dad.png' width=\"600\" height=\"600\"/>\n"
+    "![](../img/sex_mom_mom2_dad.png)\n",
+    "\n",
+    "<!-- <img src='../img/sex_mom_mom2_dad.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {
@@ -5692,7 +5698,8 @@
     "    - But grandma is irrelevant given biological mom feature\n",
     "    - Relevant features may become irrelevant given other features\n",
     "\n",
-    "<img src='../img/sex_mom_dad_grandma.png' width=\"600\" height=\"600\"/>\n"
+    "![](../img/sex_mom_dad_grandma.png)\n",
+    "<!-- <img src='../img/sex_mom_dad_grandma.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {
@@ -5711,8 +5718,9 @@
     "- General problem (\"taco Tuesday\" problem)\n",
     "    - Features can become relevant due to missing information \n",
     "\n",
+    "![](../img/sex_dad_grandma.png)\n",
     "\n",
-    "<img src='../img/sex_dad_grandma.png' width=\"600\" height=\"600\"/>\n"
+    "<!-- <img src='../img/sex_dad_grandma.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {
@@ -5731,7 +5739,9 @@
     "- General problem (Confounding)\n",
     "    - Hidden features can make irrelevant features relevant.\n",
     "\n",
-    "<img src='../img/sex_dad.png' width=\"600\" height=\"600\"/>\n"
+    "![](../img/sex_dad.png)\n",
+    "\n",
+    "<!-- <img src='../img/sex_dad.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {
@@ -5749,7 +5759,9 @@
     "- General problem (non causality)\n",
     "    - the relevant feature may not be causal \n",
     "\n",
-    "<img src='../img/sex_dad_sibling.png' width=\"600\" height=\"600\"/>\n"
+    "![](../img/sex_dad_sibling.png)\n",
+    "\n",
+    "<!-- <img src='../img/sex_dad_sibling.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {
@@ -5769,7 +5781,8 @@
     "- General problem (context specific relevance)\n",
     "    - adding a feature can make an irrelevant feature relevant\n",
     "\n",
-    "<img src='../img/sex_dad_baby.png' width=\"600\" height=\"600\"/>\n"
+    "![](../img/sex_dad_baby.png)\n",
+    "<!-- <img src='../img/sex_dad_baby.png' width=\"600\" height=\"600\"/> -->\n"
    ]
   },
   {