noshot 8.0.0__py3-none-any.whl → 10.0.0__py3-none-any.whl

noshot/data/ML TS XAI/ML/ML Lab H Sec/4. Smart House System HMM.ipynb

@@ -0,0 +1,198 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "2d6993c0-6499-43e5-bb2e-66de7de4c05c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "import networkx as nx\n",
+    "from pprint import pprint\n",
+    "import matplotlib.pyplot as plt\n",
+    "from hmmlearn.hmm import CategoricalHMM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8809eb33-e876-4f7b-b61e-9eb9c93be95e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "states = ['Kitchen', 'Bedroom', 'Living Room']\n",
+    "hidden_states = ['Cooking', 'Sleeping', 'Watching TV']"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d6570caa-6d4a-436b-802f-24d39e83b79f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "pi_states = [0.4, 0.3, 0.3]\n",
+    "pi_hidden = [0.1, 0.2, 0.7]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5f265573-be69-4218-b7af-dea81b1b80b6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "state_space = pd.Series(pi_states, index=states)\n",
+    "hidden_state_space = pd.Series(pi_hidden, index=hidden_states)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1e72ab96-f199-40c8-a074-1d7cced28bd3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "q_df = pd.DataFrame([[0.4, 0.2, 0.4], [0.45, 0.45, 0.1], [0.45, 0.25, 0.3]], \n",
+    "                    columns=states, index=states)\n",
+    "a_df = pd.DataFrame([[0.3, 0.5, 0.2], [0.1, 0.7, 0.2], [0.2, 0.3, 0.5]], \n",
+    "                    columns=hidden_states, index=hidden_states)\n",
+    "b_df = pd.DataFrame([[0.8, 0.1, 0.1], [0.1, 0.8, 0.1], [0.2, 0.1, 0.7]], \n",
+    "                    columns=states, index=hidden_states)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5cc63a43-7265-4804-b9ec-fe5c41712778",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "q_df"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d5fb306d-957d-4fcb-a0bb-9f92791e4f6f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "a_df"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "e33cb732-da75-4a7e-ba03-d15ba0b9b157",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "b_df"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "a30892b0-f44b-4c2e-b3d9-93997a3884ed",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "q = q_df.values\n",
+    "a = a_df.values\n",
+    "b = b_df.values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7f65fbf1-9b59-400b-8fe2-eedcc3153268",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def add_edges(df):\n",
+    "    return {(idx, col): df.loc[idx, col] for idx in df.index for col in df.columns}"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "67b48793-859b-4628-9f12-b3345040e788",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "edge_wts = add_edges(q_df)\n",
+    "hidden_edge_wts = add_edges(a_df)\n",
+    "emit_edge_wts = add_edges(b_df)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d9dd66a3-a331-43c6-8699-2bf81a03e600",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "G = nx.DiGraph()\n",
+    "G.add_nodes_from(states + hidden_states)\n",
+    "\n",
+    "for k, v in {**hidden_edge_wts, **emit_edge_wts}.items():\n",
+    "    G.add_edge(k[0], k[1], weight=v, label=f\"{v:.2f}\")\n",
+    "\n",
+    "pos = nx.circular_layout(G)\n",
+    "node_colors = ['skyblue' if node in hidden_states else 'lightgreen' for node in G]\n",
+    "nx.draw(G, pos, with_labels=True, arrows=True, node_color=node_colors, node_size=1000, font_size=5, font_weight='bold')\n",
+    "nx.draw_networkx_edge_labels(G, pos, edge_labels=nx.get_edge_attributes(G, 'label'), font_size=8)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "01ba9e94-fc32-4a71-aab7-35ec817dcd62",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "obs_seq = ['kitchen', 'bedroom', 'living room', 'kitchen', 'bedroom']\n",
+    "obs_map = {'kitchen': 0, 'bedroom': 1, 'living room': 2}\n",
+    "obs_idx = np.array([obs_map[o.lower()] for o in obs_seq]).reshape(-1, 1)\n",
+    "\n",
+    "model = CategoricalHMM(n_components=len(hidden_states))\n",
+    "model.startprob_ = pi_hidden\n",
+    "model.transmat_ = a\n",
+    "model.emissionprob_ = b\n",
+    "\n",
+    "logprob, path = model.decode(obs_idx, algorithm=\"viterbi\")\n",
+    "state_map = {0: 'Cooking', 1: 'Sleeping', 2: 'Watching TV'}\n",
+    "state_path = [state_map[v] for v in path]\n",
+    "\n",
+    "# Display result\n",
+    "result = pd.DataFrame({'Observation': obs_seq, 'Best_path': state_path})\n",
+    "print(result)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "NEW-VENV-1",
+   "language": "python",
+   "name": "new-venv-1"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

noshot/data/ML TS XAI/ML/ML Lab H Sec/5. Fraud Detection System HMM.ipynb

@@ -0,0 +1,201 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4be5c15a-adaa-4a0a-b378-c416ee136765",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "import networkx as network\n",
+    "import matplotlib.pyplot as plot\n",
+    "from hmmlearn.hmm import CategoricalHMM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "eebd9858-a356-448d-bc92-646a1871dd5a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "visibleStates = ['early', 'mid', 'late']\n",
+    "hiddenStates = ['Genuine User', 'Intruder']\n",
+    "\n",
+    "hiddenInitial = [0.9, 0.1]\n",
+    "visibleInitial = [0.33, 0.33, 0.33]\n",
+    "\n",
+    "hiddenTransition = pd.DataFrame([\n",
+    "    [0.7, 0.3],\n",
+    "    [0.4, 0.6]\n",
+    "], columns=hiddenStates, index=hiddenStates)\n",
+    "\n",
+    "emissionMatrix = pd.DataFrame([\n",
+    "    [0.8, 0.1, 0.1],\n",
+    "    [0.1, 0.3, 0.6]\n",
+    "], columns=visibleStates, index=hiddenStates)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d2f7f001-0f8a-4c26-a457-4319606b41d6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hiddenTransition"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "a58b0543-e154-42e5-819f-dfd70be0916b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "emissionMatrix"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c1cad5e7-a61e-4c04-98cc-920fdf763a1c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "transitionArray = hiddenTransition.values\n",
+    "emissionArray = emissionMatrix.values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "06f63c03-c3af-4183-8be6-d04e8614976c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def edgeDictionary(dataframe):\n",
+    "    return {(row, column): dataframe.loc[row, column] \n",
+    "            for row in dataframe.index \n",
+    "            for column in dataframe.columns}"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "fb957ff2-a6ad-4900-b7a3-0f87bb845d1a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "hiddenEdges = edgeDictionary(hiddenTransition)\n",
+    "emissionEdges = edgeDictionary(emissionMatrix)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "80df443e-cf00-4030-8e46-de748aeb461d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "graph = network.DiGraph()\n",
+    "graph.add_nodes_from(visibleStates + hiddenStates)\n",
+    "\n",
+    "for key, value in {**hiddenEdges, **emissionEdges}.items():\n",
+    "    graph.add_edge(key[0], key[1], weight=value, label=f\"{value:.2f}\")\n",
+    "\n",
+    "position = network.circular_layout(graph)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7dfcedd7-4298-4526-8b45-638237720b35",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "nodeColors = ['skyblue' if node in hiddenStates else 'lightgreen' for node in graph]\n",
+    "network.draw(graph, position, with_labels=True, arrows=True, node_color=nodeColors, node_size=1000, font_size=7, font_weight='bold')\n",
+    "network.draw_networkx_edge_labels(graph, position, edge_labels=network.get_edge_attributes(graph, 'label'), font_size=8)\n",
+    "plot.title(\"HMM State and Emission Graph\")\n",
+    "plot.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5bb756d0-c0d8-4509-b95b-182f30dd4a13",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "observations = ['early', 'early', 'late', 'mid', 'early', 'late']\n",
+    "observationMap = {'early': 0, 'mid': 1, 'late': 2}\n",
+    "mappedSequence = np.array([observationMap[value.lower()] for value in observations]).reshape(-1, 1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "307fdcc3-aa78-4125-a37c-67f69fa5d6ea",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = CategoricalHMM(n_components=len(hiddenStates), init_params=\"\")\n",
+    "model.startprob_ = np.array(hiddenInitial)\n",
+    "model.transmat_ = transitionArray\n",
+    "model.emissionprob_ = emissionArray\n",
+    "model.n_features = len(visibleStates)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4779218b-a7f8-4832-b5ba-b5979ba13389",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "logValue, bestPath = model.decode(mappedSequence, algorithm=\"viterbi\")\n",
+    "\n",
+    "stateMap = {index: state for index, state in enumerate(hiddenStates)}\n",
+    "decodedPath = [stateMap[state] for state in bestPath]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b89319aa-b567-443c-98a9-60198caf1882",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "resultFrame = pd.DataFrame({'Observation': observations, 'Predicted State': decodedPath})\n",
+    "print(\"\\nDecoded Path with Observations:\")\n",
+    "print(resultFrame)\n",
+    "\n",
+    "print(\"\\nLog Probability of Best Path:\", logValue)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "NEW-VENV-1",
+   "language": "python",
+   "name": "new-venv-1"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}