{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Gender Recognition by Voice Using Numerical Algorithms" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Importing Data" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import pandas as pd\n", "import seaborn as sns\n", "import numpy as np\n", "\n", "sns.set(font_scale=1.5)\n", "\n", "data = pd.read_csv(\"data/cleaned/voice_data.csv\")\n", "X_train = pd.read_csv(\"data/cleaned/X_train.csv\")\n", "y_train = pd.read_csv(\"data/cleaned/y_train.csv\")\n", "X_test = pd.read_csv(\"data/cleaned/X_test.csv\")\n", "y_test = pd.read_csv(\"data/cleaned/y_test.csv\")" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meanfreqmedianQ25Q75IQRsdlog_skewlog_kurtsp.entsfmmodemeanfunlog_minfunexp_maxfunmodindxmeandommindommaxdomdfrangelabel
0-4.049248-4.224901-2.576102-5.693607-0.2147780.4273552.9258923.177486-0.0390830.471575-2.141210-1.812038-1.3844860.636793-1.454772-1.564205-0.708404-1.431422-1.419137male
1-3.841053-3.999293-2.486885-5.588987-0.2584850.6116694.0327214.022523-0.0652360.594431-2.141210-1.079594-1.369352-0.524133-1.014103-1.561916-0.708404-1.418107-1.405818male
2-3.463066-4.095851-2.706986-3.9286990.9093261.6038484.6750894.506253-1.0837300.398261-2.141210-1.365368-1.3901310.404082-1.065344-1.563866-0.708404-1.429203-1.416917male
3-0.992157-0.759454-0.901418-0.7112050.6326900.899998-0.927599-0.8377091.5163831.797340-1.054576-1.666966-1.143909-0.5241330.614286-1.195367-0.708404-1.273867-1.261532male
4-1.530640-1.676948-1.268395-0.7920291.0055881.322561-1.055855-0.8076351.7083362.114740-0.790514-1.127233-1.2397250.1892380.289046-0.221660-0.7084040.1241540.136933male
\n", "
" ], "text/plain": [ " meanfreq median Q25 Q75 IQR sd log_skew \\\n", "0 -4.049248 -4.224901 -2.576102 -5.693607 -0.214778 0.427355 2.925892 \n", "1 -3.841053 -3.999293 -2.486885 -5.588987 -0.258485 0.611669 4.032721 \n", "2 -3.463066 -4.095851 -2.706986 -3.928699 0.909326 1.603848 4.675089 \n", "3 -0.992157 -0.759454 -0.901418 -0.711205 0.632690 0.899998 -0.927599 \n", "4 -1.530640 -1.676948 -1.268395 -0.792029 1.005588 1.322561 -1.055855 \n", "\n", " log_kurt sp.ent sfm mode meanfun log_minfun exp_maxfun \\\n", "0 3.177486 -0.039083 0.471575 -2.141210 -1.812038 -1.384486 0.636793 \n", "1 4.022523 -0.065236 0.594431 -2.141210 -1.079594 -1.369352 -0.524133 \n", "2 4.506253 -1.083730 0.398261 -2.141210 -1.365368 -1.390131 0.404082 \n", "3 -0.837709 1.516383 1.797340 -1.054576 -1.666966 -1.143909 -0.524133 \n", "4 -0.807635 1.708336 2.114740 -0.790514 -1.127233 -1.239725 0.189238 \n", "\n", " modindx meandom mindom maxdom dfrange label \n", "0 -1.454772 -1.564205 -0.708404 -1.431422 -1.419137 male \n", "1 -1.014103 -1.561916 -0.708404 -1.418107 -1.405818 male \n", "2 -1.065344 -1.563866 -0.708404 -1.429203 -1.416917 male \n", "3 0.614286 -1.195367 -0.708404 -1.273867 -1.261532 male \n", "4 0.289046 -0.221660 -0.708404 0.124154 0.136933 male " ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.head()" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 3168 entries, 0 to 3167\n", "Data columns (total 20 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 meanfreq 3168 non-null float64\n", " 1 median 3168 non-null float64\n", " 2 Q25 3168 non-null float64\n", " 3 Q75 3168 non-null float64\n", " 4 IQR 3168 non-null float64\n", " 5 sd 3168 non-null float64\n", " 6 log_skew 3168 non-null float64\n", " 7 log_kurt 3168 non-null float64\n", " 8 sp.ent 3168 non-null float64\n", " 9 sfm 3168 non-null float64\n", " 10 mode 3168 non-null float64\n", " 11 meanfun 3168 non-null float64\n", " 12 log_minfun 3168 non-null float64\n", " 13 exp_maxfun 3168 non-null float64\n", " 14 modindx 3168 non-null float64\n", " 15 meandom 3168 non-null float64\n", " 16 mindom 3168 non-null float64\n", " 17 maxdom 3168 non-null float64\n", " 18 dfrange 3168 non-null float64\n", " 19 label 3168 non-null object \n", "dtypes: float64(19), object(1)\n", "memory usage: 495.1+ KB\n" ] } ], "source": [ "data.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The meaning of the features are as follows:\n", "\n", "* `meanfreq`: mean frequency (in kHz)\n", "* `median`: median frequency (in kHz)\n", "* `Q25`: first quantile (in kHz)\n", "* `Q75`: third quantile (in kHz)\n", "* `IQR`: inter-quantile range (in kHz)\n", "* `sd`: standard deviation of frequency\n", "* `log_skew`: skewness after logarithmic transformation\n", "* `log_kurt`: kurtosis after logarithmic transformation\n", "* `sp.ent`: spectral entropy\n", "* `sfm`: spectral flatness\n", "* `mode`: mode frequency\n", "* `log_meanfun`: average of fundamental frequency measured across acoustic signal\n", "* `log_minfun`: minimum fundamental frequency measured across acoustic signal after logarithmic transformation\n", "* `exp_maxfun`: maximum fundamental frequency measured across acoustic signal after exponential transformation\n", "* `modindx`: modulation index\n", "* `meandom`: average of dominant frequency measured across acoustic signal\n", "* `mindom`: minimum of dominant frequency measured across acoustic signal\n", "* `maxdom`: maximum of dominant frequency measured across acoustic signal\n", "* `dfrange`: range of dominant frequency measured across acoustic signal\n", "* `label`: male or female" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Ensemble Vote Model\n", "\n", "Ensemble Vote Model is a technique developed by ourselves to combine multiple models in order to increase overall accuracy. We have integrated the outputs of high-performing models such as Random Forest, Support Vector Machine, and Multilayer Perception models, and selected the majority vote as the final prediction. " ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "from sklearn.neural_network import MLPClassifier\n", "from sklearn.svm import SVC\n", "from sklearn.ensemble import RandomForestClassifier\n", "\n", "\n", "class CombinedMethod(object):\n", " def __init__(self, X_train, y_train):\n", " self.X_train = X_train\n", " self.y_train = y_train\n", " self.rfClassifier = RandomForestClassifier(random_state=87).fit(\n", " X_train, y_train.squeeze()\n", " )\n", " self.svmClassifier = SVC(random_state=87, kernel=\"rbf\", gamma=0.01, C=10).fit(\n", " X_train, y_train.squeeze()\n", " )\n", " self.mlpClassifier = MLPClassifier(\n", " random_state=87, max_iter=1000, hidden_layer_sizes=(100, 100, 100, 100)\n", " ).fit(X_train, y_train.squeeze())\n", "\n", " def get_all_predictions(self, X):\n", " rf_predictions = self.rfClassifier.predict(X)\n", " svm_predictions = self.svmClassifier.predict(X)\n", " mlp_predictions = self.mlpClassifier.predict(X)\n", " return rf_predictions, svm_predictions, mlp_predictions\n", "\n", " def predict(self, X):\n", " rf_predictions = self.rfClassifier.predict(X)\n", " svm_predictions = self.svmClassifier.predict(X)\n", " mlp_predictions = self.mlpClassifier.predict(X)\n", " predictions = []\n", " for i in range(len(X)):\n", " predictions.append(\n", " np.argmax(\n", " np.bincount(\n", " [rf_predictions[i], svm_predictions[i], mlp_predictions[i]]\n", " )\n", " )\n", " )\n", " return predictions\n", "\n", " def score(self, X, y):\n", " predictions = self.predict(X)\n", " return np.sum(predictions == y.squeeze()) / len(y)\n", "\n", " def score_all(self, X, y):\n", " rf_score = self.rfClassifier.score(X, y.squeeze())\n", " svm_score = self.svmClassifier.score(X, y.squeeze())\n", " mlp_score = self.mlpClassifier.score(X, y.squeeze())\n", " combined_score = self.score(X, y)\n", " return rf_score, svm_score, mlp_score, combined_score" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We get the following results:" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train Set Scores: 1.0\n", "Test Set Scores: 0.9800664451827242\n" ] } ], "source": [ "y_train_clean = y_train == \"male\"\n", "y_test_clean = y_test == \"male\"\n", "combinedMethod = CombinedMethod(X_train, y_train_clean)\n", "print(\"Train Set Scores:\", combinedMethod.score(X_train, y_train_clean))\n", "print(\"Test Set Scores:\", combinedMethod.score(X_test, y_test_clean))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And we get the results of the three models:" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(0.9800664451827242, 0.9651162790697675, 0.973421926910299, 0.9767441860465116)" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "combinedMethod.score_all(X_test, y_test_clean)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We found that the accuracy of the Ensemble Vote model was not as ideal as we had hoped. This experience taught us the importance of carefully selecting and combining models based on their individual strengths and weaknesses, and considering the underlying assumptions and limitations of each model." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Conclusion\n", "\n", "Here are the results of the models:\n", "\n", "| Model | Training Accuracy | Testing Accuracy |\n", "| --- | --- | --- |\n", "| Classification Tree | 1.0000 | 0.9751 |\n", "| Random Forest | 1.0000 | 0.9801 |\n", "| Logistic Regression | 0.9763 | 0.9734 |\n", "| K-Nearest Neighbors | 1.0000 | 0.9817 |\n", "| Support Vector Machine | 0.9896 | 0.9834 |\n", "| Multi-Layer Perceptron | 1.0000 | 0.9734 |\n", "| Ensemble Vote | 1.0000 | 0.9800 |" ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import pandas as pd\n", "import seaborn as sns\n", "import matplotlib.pyplot as plt\n", "\n", "plt.figure(figsize=(15, 10))\n", "\n", "# Make the font bigger\n", "sns.set(font_scale=2)\n", "\n", "# Define the data\n", "data = {\n", " \"Model\": [\"CART\", \"RF\", \"LR\", \"KNN\", \"SVM\", \"MLP\", \"Ensemble\"],\n", " \"Training Accuracy\": [1.0000, 1.0000, 0.9763, 1.0000, 0.9896, 1.0000, 1.0000],\n", " \"Testing Accuracy\": [0.9751, 0.9801, 0.9734, 0.9817, 0.9834, 0.9734, 0.9800],\n", "}\n", "\n", "# Convert the data to a pandas DataFrame\n", "results = pd.DataFrame(data)\n", "\n", "# Set the style\n", "sns.set_style(\"whitegrid\")\n", "\n", "# Create the barplot\n", "ax = sns.barplot(x=\"Model\", y=\"Testing Accuracy\", data=results)\n", "\n", "# Add labels to the bars\n", "for i in range(len(results)):\n", " ax.annotate(\n", " f\"{results['Testing Accuracy'][i]:.4f}\",\n", " (i, results[\"Testing Accuracy\"][i]),\n", " ha=\"center\",\n", " va=\"bottom\",\n", " )\n", "\n", "# Add a title\n", "plt.title(\"Model Performance\")\n", "\n", "# Rotate the x-axis labels to avoid overlapping\n", "# ax.set_xticklabels(data['Model'], rotation=45)\n", "\n", "# Show the plot\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.12" } }, "nbformat": 4, "nbformat_minor": 1 }