{"id":9408,"date":"2024-03-05T06:00:43","date_gmt":"2024-03-05T14:00:43","guid":{"rendered":"https:\/\/live-cometml.pantheonsite.io\/?p=9408"},"modified":"2025-04-24T17:03:02","modified_gmt":"2025-04-24T17:03:02","slug":"pima-indian-diabetes-prediction","status":"publish","type":"post","link":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\/","title":{"rendered":"Pima Indian Diabetes Prediction"},"content":{"rendered":"\n
\"\"\/<\/figure>\n\n\n\n

This project aims to analyze the medical factors of a patient, such as Glucose Level, Blood Pressure, Skin Thickness, Insulin Level, and many others, to predict whether the patient has diabetes or not.<\/p>\n\n\n\n

About the Dataset<\/h4>\n\n\n\n

This dataset is originally from the National Institute of Diabetes and Digestive and Kidney Diseases. The objective of the dataset is to diagnostically predict whether or not a patient has diabetes based on specific diagnostic measurements included in the dataset. Several constraints were placed on selecting these instances from a larger database. In particular, all patients here are females at least 21 years old of Pima Indian heritage.<\/p>\n\n\n\n

The dataset consists of several medical predictor variables and one target variable, Outcome. Predictor variables include the number of pregnancies the patient has had, their BMI, insulin level, age, etc.<\/p>\n\n\n\n

Data Dictionary<\/h4>\n\n\n\n
\"\"\/
Figure 1: Data Dictionary<\/figcaption><\/figure>\n\n\n\n

 <\/p>\n\n\n\n

Importing the libraries<\/p>\n\n\n\n

import<\/span> pandas as<\/span> pd\nimport<\/span> numpy as<\/span> np\nimport<\/span> matplotlib.pyplot as<\/span> plt\nimport<\/span> seaborn as<\/span> sns<\/span><\/pre>\n\n\n\n
Loading the dataset<\/p>\n\n\n\n
df = pd.read_csv(\"diabetes.csv\"<\/span>)\ndf.head()<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 2: Dataset<\/figcaption><\/figure>\n\n\n\n
Data Preprocessing<\/h3>\n\n\n\n
Checking the shape of the dataset<\/p>\n\n\n\n
df.shape\n=> (768<\/span>, 9<\/span>)<\/span><\/pre>\n\n\n\n
Unique values in the dataset<\/p>\n\n\n\n
variables = ['Pregnancies'<\/span>,'Glucose'<\/span>,'BloodPressure'<\/span>,'SkinThickness'<\/span>,'Insulin'<\/span>,'BMI'<\/span>,'DiabetesPedigreeFunction'<\/span>,'Age'<\/span>,'Outcome'<\/span>]\nfor<\/span> i in<\/span> variables:\n    print<\/span>(df[i].unique())<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 3: Unique Values<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
In the dataset, the variables except Pregnancies and Outcome cannot have a value of 0 because it is impossible to have 0 Glucose Levels or 0 Blood Pressure. So, this will be counted as incorrect information.<\/p>\n\n\n\n
Checking the count of value 0 in the variables<\/p>\n\n\n\n
variables = ['Glucose'<\/span>,'BloodPressure'<\/span>,'SkinThickness'<\/span>,'Insulin'<\/span>,'BMI'<\/span>,'DiabetesPedigreeFunction'<\/span>,'Age'<\/span>,]\nfor<\/span> i in<\/span> variables:\n    c = 0<\/span>\n    for<\/span> x in<\/span> (df[i]):\n        if<\/span> x == 0<\/span>:\n            c = c + 1<\/span>\n    print<\/span>(i,c)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 4: Zero value count<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Now that I have a count of incorrect values in the variables, I will be replacing these values.<\/p>\n\n\n\n
#replacing the missing values with the mean<\/span>\nvariables = ['Glucose'<\/span>,'BloodPressure'<\/span>,'SkinThickness'<\/span>,'Insulin'<\/span>,'BMI'<\/span>]\nfor<\/span> i in<\/span> variables:\n    df[i].replace(0<\/span>,df[i].mean(),inplace=True<\/span>)<\/span><\/pre>\n\n\n\n
#checking to make sure that incorrect values are replace<\/span>\nfor<\/span> i in<\/span> variables:\n    c = 0<\/span>\n    for<\/span> x in<\/span> (df[i]):\n        if<\/span> x == 0<\/span>:\n            c = c + 1<\/span>\n    print<\/span>(i,c)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 5: Zero values replaced<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Checking for missing values<\/p>\n\n\n\n
df.info()<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 6: Null values<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Descriptive Statistics<\/p>\n\n\n\n
df.describe()<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 7: Descriptive Statistics<\/figcaption><\/figure>\n\n\n\n
df.head()<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 8: Dataset<\/figcaption><\/figure>\n\n\n\n
Exploratory Data Analysis<\/h3>\n\n\n\n
In the exploratory data analysis, I will look at the data distribution, the correlation between the features, and the relationship between the features and the target variable. I will start by looking at the data distribution, followed by the relationship between the target variable and independent variables.<\/p>\n\n\n\n
Diabetes Count<\/h4>\n\n\n\n
plt.figure(figsize=(5<\/span>,5<\/span>))\nplt.pie(df['Outcome'<\/span>].value_counts(), labels=['No Diabetes'<\/span>, 'Diabetes'<\/span>], autopct='%1.1f%%'<\/span>, shadow=False<\/span>, startangle=90<\/span>)\nplt.title('Diabetes Outcome'<\/span>)\nplt.show()<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 9: Diabetes count<\/figcaption><\/figure>\n\n\n\n
Age Distribution and Diabetes<\/h4>\n\n\n\n
sns.catplot(x=\"Outcome\"<\/span>, y=\"Age\"<\/span>, kind=\"swarm\"<\/span>, data=df)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 10: Age and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
From the graph, it is quite clear that most patients are adults aged 20\u201330 years. Patients in the age range 40\u201355 years are more prone to diabetes, as compared to other age groups. Since the number of adults in the age group 20\u201330 years is greater, the number of patients with diabetes is also more as compared to other age groups.<\/p>\n\n\n\n
Pregnancies and Diabetes<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>,y='Pregnancies'<\/span>,data=df,ax=ax[0<\/span>])\nsns.violinplot(x='Outcome'<\/span>,y='Pregnancies'<\/span>,data=df,ax=ax[1<\/span>])<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 11: Pregnancies and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
The boxplot and violin plot show a strange relationship between the number of pregnancies and diabetes. According to the graphs, the increased number of pregnancies highlights an increased risk of diabetes.<\/p>\n\n\n\n
Glucose and Diabetes<\/h4>\n\n\n\n
sns.boxplot(x='Outcome'<\/span>, y='Glucose'<\/span>, data=df).set_title('Glucose vs Diabetes'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 12: Glucose and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Glucose level plays a significant role in determining whether the patient has diabetes. Patients with a median glucose level of less than 120 are more likely to be nondiabetic. Patients with a median glucose level greater than 140 are more likely to be diabetic. Therefore, high glucose levels are a good indicator of diabetes.<\/p>\n\n\n\n
Blood Pressure and Diabetes<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>, y='BloodPressure'<\/span>, data=df, ax=ax[0<\/span>]).set_title('BloodPressure vs Diabetes'<\/span>)\nsns.violinplot(x='Outcome'<\/span>, y='BloodPressure'<\/span>, data=df, ax=ax[1<\/span>]).set_title('BloodPressure vs Diabetes'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 13: Blood Pressure and Glucose<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Both the boxplot and violin plot provide a clear understanding of the relationship between blood pressure and diabetes. The boxplot shows that the median blood pressure for diabetic patients is slightly higher than nondiabetic patients. The violin plot shows that the distribution of blood pressure for diabetic patients is slightly higher than for nondiabetic patients. However, there has not been enough evidence to conclude that blood pressure is a good predictor of diabetes.<\/p>\n\n\n\n
Skin Thickness and Diabetes<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>, y='SkinThickness'<\/span>, data=df,ax=ax[0<\/span>]).set_title('SkinThickness vs Diabetes'<\/span>)\nsns.violinplot(x='Outcome'<\/span>, y='SkinThickness'<\/span>, data=df,ax=ax[1<\/span>]).set_title('SkinThickness vs Diabetes'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 14: Skin Thickness and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Here, both the boxplot and violinplot reveal the effect of diabetes on skin thickness. As observed in the boxplot, the median skin thickness is higher for diabetic patients than nondiabetic patients. Nondiabetic patients have a median skin thickness of nearly 20, compared to almost 30 in diabetic patients. The violin plot shows the distribution of patients’ skin thickness among the patients, where the nondiabetic ones have a greater distribution near 20, people with diabetes have a smaller distribution near 20, and increased distribution near 30. Therefore, skin thickness can be an indicator of diabetes.<\/p>\n\n\n\n
Insulin and Diabetes<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>,y='Insulin'<\/span>,data=df,ax=ax[0<\/span>]).set_title('Insulin vs Diabetes'<\/span>)\nsns.violinplot(x='Outcome'<\/span>,y='Insulin'<\/span>,data=df,ax=ax[1<\/span>]).set_title('Insulin vs Diabetes'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 15: Insulin and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Insulin is a major body hormone that regulates glucose metabolism. It’s required for the body to use sugars, fats, and proteins efficiently. Any change in insulin amount in the body would also result in a change in glucose levels. Here, the boxplot and violinplot show the distribution of insulin levels in patients. In nondiabetic patients, the insulin level is near 100, whereas in diabetic patients, the insulin level is near 200. In the violin plot, we can see that the distribution of insulin levels in nondiabetic patients is more spread out near 100, whereas, in diabetic patients, the distribution is contracted and shows a little spread in higher insulin levels. This indicates that the insulin level is a good indicator of diabetes.<\/p>\n\n\n\n
BMI and Diabetes<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>,y='BMI'<\/span>,data=df,ax=ax[0<\/span>])\nsns.violinplot(x='Outcome'<\/span>,y='BMI'<\/span>,data=df,ax=ax[1<\/span>])<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 16: BMI and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
Both graphs highlight the role of BMI in diabetes prediction. Nondiabetic patients have a normal BMI within the range of 25\u201335, whereas diabetic patients have a BMI greater than 35. The violin plot reveals the BMI distribution, where the nondiabetic patients have an increased spread from 25 to 35, with narrows after 35. However, in diabetic patients, there is an increased spread at 35 and an increased spread at 45\u201350 compared to nondiabetic patients. Therefore, BMI is a good predictor of diabetes, and obese people are more likely to be diabetic.<\/p>\n\n\n\n
Diabetes Pedigree Function and Diabetes Outcome<\/h4>\n\n\n\n
fig,ax = plt.subplots(1<\/span>,2<\/span>,figsize=(15<\/span>,5<\/span>))\nsns.boxplot(x='Outcome'<\/span>,y='DiabetesPedigreeFunction'<\/span>,data=df,ax=ax[0<\/span>]).set_title('Diabetes Pedigree Function'<\/span>)\nsns.violinplot(x='Outcome'<\/span>,y='DiabetesPedigreeFunction'<\/span>,data=df,ax=ax[1<\/span>]).set_title('Diabetes Pedigree Function'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 17: Diabetes Pedigree Function and Diabetes<\/figcaption><\/figure>\n\n\n\n
 <\/p>\n\n\n\n
The Diabetes Pedigree Function (DPF) calculates diabetes likelihood depending on the subject’s age and diabetic family history. The boxplot shows that patients with lower DPF are much less likely to have diabetes. The patients with higher DPF are much more likely to have diabetes. In the violin plot, the majority of the nondiabetic patients have a DPF of 0.25\u20130.35, whereas the diabetic patients have an increased DPF, which is shown by their distribution in the violin plot where there is an increased spread in the DPF from 0.5 -1.5. Therefore, the DPF is a good indicator of diabetes.<\/p>\n\n\n\n
Correlation Matrix Heatmap<\/h3>\n\n\n\n
#correlation heatmap<\/span>\nplt.figure(figsize=(12<\/span>,12<\/span>))\nsns.heatmap(df.corr(), annot=True<\/span>, cmap='coolwarm'<\/span>).set_title('Correlation Heatmap'<\/span>)<\/span><\/pre>\n\n\n\n
\"\"\/
Figure 18: Correlation Matrix Heatmap<\/figcaption><\/figure>\n\n\n\n
Train Test Split<\/h3>\n\n\n\n
from<\/span> sklearn.model_selection import<\/span> train_test_split\nX_train, X_test, y_train, y_test = train_test_split(df.drop('Outcome'<\/span>,axis=1<\/span>),df['Outcome'<\/span>],test_size=0.2<\/span>,random_state=42<\/span>)<\/span><\/pre>\n\n\n\n
Diabetes Prediction<\/h3>\n\n\n\n
For predicting diabetes, I will be using the following algorithms:<\/p>\n\n\n\n
    \n
  1. Logistic Regression<\/li>\n\n\n\n
  2. Random Forest Classifier<\/li>\n\n\n\n
  3. Support Vector Machine<\/li>\n<\/ol>\n\n\n\n
    Logistic Regression<\/h4>\n\n\n\n
    #building model<\/span>\nfrom<\/span> sklearn.linear_model import<\/span> LogisticRegression\nlr = LogisticRegression()\n\n#training the model<\/span>\nlr.fit(X_train,y_train)\n\n#training accuracy<\/span>\nlr.score(X_train,y_train)\n=> 0.7719869706840391<\/span>\n\n#predicted outcomes<\/span>\nlr_pred = lr.predict(X_test)<\/span><\/pre>\n\n\n\n
    Random Forest Classifier<\/h4>\n\n\n\n
    #buidling model<\/span>\nfrom<\/span> sklearn.ensemble import<\/span> RandomForestClassifier\nrfc = RandomForestClassifier(n_estimators=100<\/span>,random_state=42<\/span>)\n\n#training model<\/span>\nrfc.fit(X_train, y_train)\n\n#training accuracy<\/span>\nrfc.score(X_train, y_train)\n=> 0.978543940<\/span>\n\n#predicted outcomes<\/span>\nrfc_pred = rfc.predict(X_test)<\/span><\/pre>\n\n\n\n
    Support Vector Machine (SVM)<\/h4>\n\n\n\n
    #building model<\/span>\nfrom<\/span> sklearn.svm import<\/span> SVC\nsvm = SVC(kernel='linear'<\/span>, random_state=0<\/span>)\n\n#training the model<\/span>\nsvm.fit(X_train, y_train)\n\n#training the model<\/span>\nsvm.score(X_test, y_test)\n=> 0.7597402597402597<\/span>\n\n#predicting outcomes<\/span>\nsvm_pred = svm.predict(X_test)<\/span><\/pre>\n\n\n\n
    Model Evaluation<\/h3>\n\n\n\n
    Evaluating the Logistic Regression Model<\/h4>\n\n\n\n
    Confusion Matrix Heatmap<\/h4>\n\n\n\n
    from<\/span> sklearn.metrics import<\/span> confusion_matrix\nsns.heatmap(confusion_matrix(y_test, lr_pred), annot=True<\/span>, cmap='Blues'<\/span>)\nplt.xlabel('Predicted Values'<\/span>)\nplt.ylabel('Actual Values'<\/span>)\nplt.title('Confusion Matrix for Logistic Regression'<\/span>)\nplt.show()<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 19: Logistic Regression Confusion Matrix<\/figcaption><\/figure>\n\n\n\n
    Distribution Plot<\/h4>\n\n\n\n
    ax = sns.distplot(y_test, color='r'<\/span>,  label='Actual Value'<\/span>,hist=False<\/span>)\nsns.distplot(lr_pred, color='b'<\/span>, label='Predicted Value'<\/span>,hist=False<\/span>,ax=ax)\nplt.title('Actual vs Predicted Value Logistic Regression'<\/span>)\nplt.xlabel('Outcome'<\/span>)\nplt.ylabel('Count'<\/span>)<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 20: Distribution Plot Logistic Regression<\/figcaption><\/figure>\n\n\n\n
    Classification Report<\/h4>\n\n\n\n
    from<\/span> sklearn.metrics import<\/span> classification_report\nprint<\/span>(classification_report(y_test, lr_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 21: Classification Report Logistic Regression<\/figcaption><\/figure>\n\n\n\n
    from<\/span> sklearn.metrics import<\/span> accuracy_score,mean_absolute_error,mean_squared_error,r2_score\nprint<\/span>('Accuracy Score: '<\/span>,accuracy_score(y_test,lr_pred))\nprint<\/span>('Mean Absolute Error: '<\/span>,mean_absolute_error(y_test,lr_pred))\nprint<\/span>('Mean Squared Error: '<\/span>,mean_squared_error(y_test,lr_pred))\nprint<\/span>('R2 Score: '<\/span>,r2_score(y_test,lr_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 22: Logistic Regression Model Metrics<\/figcaption><\/figure>\n\n\n\n
    Evaluating Random Forest Classifier<\/h4>\n\n\n\n
    Confusion Matrix Heatmap<\/h4>\n\n\n\n
    sns.heatmap(confusion_matrix(y_test, rfc_pred), annot=True<\/span>, cmap='Blues'<\/span>)\nplt.xlabel('Predicted Values'<\/span>)\nplt.ylabel('Actual Values'<\/span>)\nplt.title('Confusion Matrix for Logistic Regression'<\/span>)\nplt.show()<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 23: Confusion Matrix Random Forest<\/figcaption><\/figure>\n\n\n\n
    Distribution Plot<\/h4>\n\n\n\n
    ax = sns.distplot(y_test, color='r'<\/span>,  label='Actual Value'<\/span>,hist=False<\/span>)\nsns.distplot(rfc_pred, color='b'<\/span>, label='Predicted Value'<\/span>,hist=False<\/span>,ax=ax)\nplt.title('Actual vs Predicted Value Logistic Regression'<\/span>)\nplt.xlabel('Outcome'<\/span>)\nplt.ylabel('Count'<\/span>)<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 24: Distribution Plot Random Forest<\/figcaption><\/figure>\n\n\n\n
    Classification Report<\/h4>\n\n\n\n
    print<\/span>(classification_report(y_test, rfc_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 25: Classification Report Random Forest<\/figcaption><\/figure>\n\n\n\n
    print<\/span>('Accuracy Score: '<\/span>,accuracy_score(y_test,rfc_pred))\nprint<\/span>('Mean Absolute Error: '<\/span>,mean_absolute_error(y_test,rfc_pred))\nprint<\/span>('Mean Squared Error: '<\/span>,mean_squared_error(y_test,rfc_pred))\nprint<\/span>('R2 Score: '<\/span>,r2_score(y_test,rfc_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 26: Model Metrics Random Forest<\/figcaption><\/figure>\n\n\n\n
    Evaluating SVM Model<\/h4>\n\n\n\n
    Confusion Matrix Heatmap<\/h4>\n\n\n\n
    sns.heatmap(confusion_matrix(y_test, svm_pred), annot=True<\/span>, cmap='Blues'<\/span>)\nplt.xlabel('Predicted Values'<\/span>)\nplt.ylabel('Actual Values'<\/span>)\nplt.title('Confusion Matrix for Logistic Regression'<\/span>)\nplt.show()<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 27: Confusion Matrix SVM<\/figcaption><\/figure>\n\n\n\n
    Distribution Plot<\/h4>\n\n\n\n
    ax = sns.distplot(y_test, color='r'<\/span>,  label='Actual Value'<\/span>,hist=False<\/span>)\nsns.distplot(svm_pred, color='b'<\/span>, label='Predicted Value'<\/span>,hist=False<\/span>,ax=ax)\nplt.title('Actual vs Predicted Value Logistic Regression'<\/span>)\nplt.xlabel('Outcome'<\/span>)\nplt.ylabel('Count'<\/span>)<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 28: Distribution Plot SVM<\/figcaption><\/figure>\n\n\n\n
    Classification Report<\/h4>\n\n\n\n
    print<\/span>(classification_report(y_test, rfc_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 29: Classification Report SVM<\/figcaption><\/figure>\n\n\n\n
    print<\/span>('Accuracy Score: '<\/span>,accuracy_score(y_test,svm_pred))\nprint<\/span>('Mean Absolute Error: '<\/span>,mean_absolute_error(y_test,svm_pred))\nprint<\/span>('Mean Squared Error: '<\/span>,mean_squared_error(y_test,svm_pred))\nprint<\/span>('R2 Score: '<\/span>,r2_score(y_test,svm_pred))<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 30: Model Metrics SVM<\/figcaption><\/figure>\n\n\n\n
    Comparing the Models<\/h3>\n\n\n\n
    #comparing the accuracy of different models<\/span>\nsns.barplot(x=['Logistic Regression'<\/span>, 'RandomForestClassifier'<\/span>, 'SVM'<\/span>], y=[0.7792207792207793,0.7662337662337663,0.7597402597402597])\nplt.xlabel('Classifier Models'<\/span>)\nplt.ylabel('Accuracy'<\/span>)\nplt.title('Comparison of different models'<\/span>)<\/span><\/pre>\n\n\n\n
    \"\"\/
    Figure 31: Model Comparison<\/figcaption><\/figure>\n\n\n\n
    Conclusion<\/h3>\n\n\n\n
    From the exploratory data analysis, I have concluded that the risk of diabetes depends upon the following factors:<\/p>\n\n\n\n
      \n
    1. Glucose level<\/li>\n\n\n\n
    2. Number of pregnancies<\/li>\n\n\n\n
    3. Skin Thickness<\/li>\n\n\n\n
    4. Insulin level<\/li>\n\n\n\n
    5. BMI<\/li>\n<\/ol>\n\n\n\n
      With an increase in Glucose level, insulin level, BMI, and number of pregnancies, the risk of diabetes increases. However, the number of pregnancies has a strange effect on the risk of diabetes, which the data couldn’t explain. The risk of diabetes also increases with an increase in skin thickness.<\/p>\n\n\n\n
      Coming to the classification models, Logistic Regression outperformed Random Forest and SVM with 78% accuracy. The model’s accuracy can be improved by increasing the size of the dataset. The dataset used for this project was very small and had only 768 rows.<\/p>\n","protected":false},"excerpt":{"rendered":"
      This project aims to analyze the medical factors of a patient, such as Glucose Level, Blood Pressure, Skin Thickness, Insulin Level, and many others, to predict whether the patient has diabetes or not. About the Dataset This dataset is originally from the National Institute of Diabetes and Digestive and Kidney Diseases. The objective of the dataset […]<\/p>\n","protected":false},"author":120,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"customer_name":"","customer_description":"","customer_industry":"","customer_technologies":"","customer_logo":"","footnotes":""},"categories":[7],"tags":[],"coauthors":[217],"class_list":["post-9408","post","type-post","status-publish","format-standard","hentry","category-tutorials"],"yoast_head":"\nPima Indian Diabetes Prediction - Comet<\/title>\n<meta name=\"description\" content=\"This project covers diabetes prediction by looking at different medical factors, such as glucose level, blood pressure, etc.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Pima Indian Diabetes Prediction\" \/>\n<meta property=\"og:description\" content=\"This project covers diabetes prediction by looking at different medical factors, such as glucose level, blood pressure, etc.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\" \/>\n<meta property=\"og:site_name\" content=\"Comet\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/cometdotml\" \/>\n<meta property=\"article:published_time\" content=\"2024-03-05T14:00:43+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-04-24T17:03:02+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg\" \/>\n<meta name=\"author\" content=\"Sukhman Singh\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@Cometml\" \/>\n<meta name=\"twitter:site\" content=\"@Cometml\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Sukhman Singh\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"15 minutes\" \/>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Pima Indian Diabetes Prediction - Comet","description":"This project covers diabetes prediction by looking at different medical factors, such as glucose level, blood pressure, etc.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction","og_locale":"en_US","og_type":"article","og_title":"Pima Indian Diabetes Prediction","og_description":"This project covers diabetes prediction by looking at different medical factors, such as glucose level, blood pressure, etc.","og_url":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction","og_site_name":"Comet","article_publisher":"https:\/\/www.facebook.com\/cometdotml","article_published_time":"2024-03-05T14:00:43+00:00","article_modified_time":"2025-04-24T17:03:02+00:00","og_image":[{"url":"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg","type":"","width":"","height":""}],"author":"Sukhman Singh","twitter_card":"summary_large_image","twitter_creator":"@Cometml","twitter_site":"@Cometml","twitter_misc":{"Written by":"Sukhman Singh","Est. reading time":"15 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#article","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\/"},"author":{"name":"Sukhman Singh","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/03fee0ceb719563b09dcf21796f85455"},"headline":"Pima Indian Diabetes Prediction","datePublished":"2024-03-05T14:00:43+00:00","dateModified":"2025-04-24T17:03:02+00:00","mainEntityOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\/"},"wordCount":1284,"publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#primaryimage"},"thumbnailUrl":"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg","articleSection":["Tutorials"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction\/","url":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction","name":"Pima Indian Diabetes Prediction - Comet","isPartOf":{"@id":"https:\/\/www.comet.com\/site\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#primaryimage"},"image":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#primaryimage"},"thumbnailUrl":"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg","datePublished":"2024-03-05T14:00:43+00:00","dateModified":"2025-04-24T17:03:02+00:00","description":"This project covers diabetes prediction by looking at different medical factors, such as glucose level, blood pressure, etc.","breadcrumb":{"@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#primaryimage","url":"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg","contentUrl":"https:\/\/cdn-images-1.medium.com\/max\/800\/0*7p7zUmwZ02iokYmP.jpg"},{"@type":"BreadcrumbList","@id":"https:\/\/www.comet.com\/site\/blog\/pima-indian-diabetes-prediction#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.comet.com\/site\/"},{"@type":"ListItem","position":2,"name":"Pima Indian Diabetes Prediction"}]},{"@type":"WebSite","@id":"https:\/\/www.comet.com\/site\/#website","url":"https:\/\/www.comet.com\/site\/","name":"Comet","description":"Build Better Models Faster","publisher":{"@id":"https:\/\/www.comet.com\/site\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.comet.com\/site\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.comet.com\/site\/#organization","name":"Comet ML, Inc.","alternateName":"Comet","url":"https:\/\/www.comet.com\/site\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/","url":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","contentUrl":"https:\/\/www.comet.com\/site\/wp-content\/uploads\/2025\/01\/logo_comet_square.png","width":310,"height":310,"caption":"Comet ML, Inc."},"image":{"@id":"https:\/\/www.comet.com\/site\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/cometdotml","https:\/\/x.com\/Cometml","https:\/\/www.youtube.com\/channel\/UCmN63HKvfXSCS-UwVwmK8Hw"]},{"@type":"Person","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/03fee0ceb719563b09dcf21796f85455","name":"Sukhman Singh","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.comet.com\/site\/#\/schema\/person\/image\/cbb8e2cab1e32e43711b4e44c7711e76","url":"https:\/\/secure.gravatar.com\/avatar\/2e74502dcaa20f8fe126f4e1eef424b9227d93803c9cc857de010fbca0b41265?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/2e74502dcaa20f8fe126f4e1eef424b9227d93803c9cc857de010fbca0b41265?s=96&d=mm&r=g","caption":"Sukhman Singh"},"url":"https:\/\/www.comet.com\/site\/blog\/author\/sukhmansinghbhogalgmail-com\/"}]}},"_links":{"self":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/9408","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/users\/120"}],"replies":[{"embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/comments?post=9408"}],"version-history":[{"count":1,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/9408\/revisions"}],"predecessor-version":[{"id":15378,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/posts\/9408\/revisions\/15378"}],"wp:attachment":[{"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/media?parent=9408"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/categories?post=9408"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/tags?post=9408"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.comet.com\/site\/wp-json\/wp\/v2\/coauthors?post=9408"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}