We can clearly see\nthat the correlation between Ma1 to Ma 6 are close and Exu1-Exu8 are close to\neach other, then both ma\u2019s and exu\u2019s are correlated than the others. Features\nlike- amfm, abn, quality do not correlate with any of the other features.<\/p>\n\n\n\n
Standardisation<\/strong>: this is an\nimportant step before modelling as the features are needed to be in particular\nrange. Consider this example, if feature age 50 and insulin 0.5 is fed into ML\nalgorithm the machine thinks that age is more important as it has higher value.\nFor this reason we need to normalize the dataset leaving out the target column.\nAfter completing this standardisation process the dataset looks like this- <\/p>\n\n\n\nWe are doing this before visuals and modeling so that the features get normalize, this is achieved by using Standard scalar. Below are the few lines of normalized features \u2013<\/p>\n\n\n\n
Features like Euclidean distance and Diameter of the optic disc seems like important features . <\/p>\n\n\n\n
Features like Ma1 \u2013 Ma6 are numerical and have been normalized. These MicroAneurisms features are respectively dealt with the alpha rates, which start from 0.5 to 1 .<\/p>\n\n\n\n
Features like EXu1 \u2013 Exu8 also have been normalized; they are represented by a set of points which represents lesions in the retina.<\/p>\n\n\n\n
Principal Component Analysis: PCA is essentially a method that reduces the dimension of the feature space in such a way that new variables are orthogonal to each other (i.e. they are independent or not correlated). Here we are going to view the dataset with 2 principal components to see if we can split the dataset easily with basic ml algorithms.<\/strong><\/p>\n\n\n\nAs\nthe data points overlap, this might be difficult to separate linearly. Let us\ntry to use various machine learning algorithms to predict diabetic retinopathy\nin a patient.<\/strong><\/strong><\/p>\n\n\n\nPredictive modelling:<\/strong><\/p>\n\n\n\nWe will be\nusing different machine learning algorithms such as SVM(Support Vector\nMachine), Logistic Regression , Random Forest classifier, Decision tree, KNN(K-\nNearest Neighbours) and MLP(Multilayer Perceptron).<\/p>\n\n\n\n
SVM: <\/strong>we will be using support vector machine from sklearn python package. There are two kernels in svm, namely Linear and rbf, and we will be testing both the algorithms.<\/p>\n\n\n\nLogistic Regression: <\/strong>this algorithm is also from sklearn python package, here the main hyperparameters are C and Penalty(L1 or L2)<\/p>\n\n\n\nRandom Forest<\/strong>: this algorithm is a type of ensemble model which works well for classification problems. Let us first find the important features and then predict diabetes.<\/p>\n\n\n\nFeatures like amfm, abn and quality seem like the lowest of all and too low of importance shown than others. This might decide the outcome of the accuracy later, first we will try running with all the features.<\/p>\n\n\n\n
KNN: <\/strong>K- nearest neighbour is a machine learning algorithm which uses distance\nmetrics to find the closest neighbours of our features, we need to find the\nvalue of k to get the best accuracy at a particular value.<\/p>\n\n\n\nWe will be testing with\nmany numbers of k- neighbours so that we can find the best amount for k to\nachieve better accuracy.<\/p>\n\n\n\n
For this dataset let\u2019s set the value of k as 20.<\/p>\n\n\n\n
Accuracy does not\nseem to get higher than 66%, so we will stop with 20 nearest neighbours and 20\nth value for k got 65%.<\/p>\n\n\n\n
Adaboost classifier<\/strong>: Let us now try Ensemble model (which takes Decision tree classifier as its base algorithm and same hyperparameters as decision tree is set)<\/p>\n\n\n\nNext, we will be trying MULTI-LAYER PERCEPTRON<\/strong> -with different iteration of different layers and learning rate.<\/p>\n\n\n\n After a long trial of testing learning\nrate(0.1,0.01,0.001,0.0001), solver: adam, lgbfs, and choosing different layers\n, the best accuracy is got from \u2013 <\/p>\n\n\n\n
Learning rate\n\u2013 1<\/p>\n\n\n\n
Solver \u2013 Lgbfs<\/p>\n\n\n\n
Hidden layer\nsize \u2013 7<\/p>\n\n\n\n
Random state \u2013\n1<\/p>\n\n\n\n
We got an\naccuracy of 73%, which is better than most of the basic algorithms.<\/p>\n\n\n\n
Cross-validation<\/strong>:<\/strong> Cross-validation is a\nresampling procedure used to evaluate machine learning models on a limited data\nsample.<\/p>\n\n\n\nThe general\nprocedure is as follows:<\/p>\n\n\n\n
Shuffle the dataset randomly.<\/li> Split the dataset into k groups.<\/li> For each unique group:<\/li><\/ol>\n\n\n\nTake the\n group as a holdout or test data set.<\/li> Take the\n remaining groups as a training data set.<\/li> Fit a model\n on the training set and evaluate it on the test set.<\/li> Retain the\n evaluation score and discard the model.<\/li> Summarize the skill of the model using the sample\nof model evaluation scores.<\/li><\/ol>\n\n\n\nNow we will be using this cross-validation for our algorithms and check which gives us better accuracy \u2013 let K be 10<\/p>\n\n\n\n
Multi-layer perceptron has done well in accuracy while using cross-validation with 75%.<\/p>\n\n\n\n
Before we would have seen that some of the features are not so important than the others. These features are amfm , quality and abn, lets delete these out and test again with mlp model.<\/p>\n\n\n\n
After training,\nwe can see that this model with deleting unwanted features had improved a lot\nbetter than before with more than77% accuracy.<\/p>\n\n\n\n
Classification report \u2013 <\/p>\n\n\n\n
AUC \u2013 ROC\ncurve: <\/strong><\/p>\n\n\n\nAUC\u2013ROC curve is the model selection metric for bi\u2013multi-class classification problem. ROC is a probability curve for different classes. ROC tells us how good the model is for distinguishing the given classes, in terms of the predicted probability. A typical ROC curve has False Positive Rate (FPR) on the X-axis and True Positive Rate (TPR) on the Y-axis. The area covered by the curve is the area between the orange line (ROC) and the axis. This area covered is AUC. The bigger the area covered, the better the machine learning models is at distinguishing the given classes. The ideal value for AUC is 1.<\/p>\n\n\n\n
Further Proceedings:<\/strong><\/p>\n\n\n\nThe dataset contains the details of Messidor dataset image details, and\nwe can achieve better accuracy with the help of scanned images of retina to\nmake better impact inaccuracy.As this dataset is so crucial and important in the science industry,\nthere should be more contributors to the dataset, which leads in better ml\nmodel.<\/li><\/ul>With Image dataset of Messidor, we can use various CNN, Inception V3 and\nother better algorithms to achieve better accuracy.<\/li><\/ul><\/li><\/ul>\n\n\n\nWhy Guires<\/h3>\n\n\n\n Guires\nData analytics mission is to democratize AI for healthcare industries. The team\nof data science expert use the power of AI to solve business and social\nchallenges. We are a pioneer in the\nresearch field for more than fifteen years and offer end to end solution for\nthe firm to set the direction for the company and support analytical frameworks\nfor better understanding and making strategic decisions. We provide appropriate\nsolutions using your existing volume of data available in varying degree of\ncomplexities that cannot be processed using traditional technologies,\nprocessing methods, or any commercial off the shelf solutions. By outsourcing\nbig data to us, we can analyze events that have happened within and outside an\norganization and correlate those to provide near accurate insights into what\ndrove the outcome. Our big data analytics solutions are fast, scalable and\npossess flexible processing.<\/p>\n\n\n\n
We use powerful algorithms, business rules, and statistical models. We work with text, image, audio, video and\nmachine data. Our medical experts understand the different layers of data being\nintegrated and what granularity levels of integration can be completed to\ncreate the holistic picture. Our team creates the foundational structure for\nanalytics and visualization of the data. Our data analytics team is well\nequipped with advanced mathematical degrees, statisticians with multiple\nspecialist degrees who can apply cutting-edge data mining techniques thereby\nenabling our clients to gain rich insights into existing customers and unearth\nhigh potential prospects.<\/p>\n\n\n\n
How\ncan you make the most of predictive analytics? Let us help you get started.<\/p>\n\n\n\n
Get predictive\nanalytics working for you. Contact Guires expert.<\/p>\n","protected":false},"excerpt":{"rendered":"
The Challenges Diabetic retinopathy, also known as diabetic eye disease, is a medical condition in which damage occurs to the retina due to diabetes mellitus. It is a leading cause of blindness. Diabetic retinopathy affects up to 80% of those who have had diabetes for 20 years or more. At least 90% of new cases could be reduced with proper treatment […]<\/p>\n","protected":false},"author":1,"featured_media":709,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[55],"tags":[],"class_list":["post-671","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-use-cases"],"_links":{"self":[{"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/posts\/671","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/comments?post=671"}],"version-history":[{"count":2,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/posts\/671\/revisions"}],"predecessor-version":[{"id":710,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/posts\/671\/revisions\/710"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/media\/709"}],"wp:attachment":[{"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/media?parent=671"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/categories?post=671"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/guires.uk\/newsroom\/wp-json\/wp\/v2\/tags?post=671"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-96.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-97.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-98.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-99.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-100.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-101.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-102.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-103.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-104.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-105.png\")
<\/figure>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-106.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-107.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-108.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-109.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-110.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-111.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-112.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-113.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-114.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-115.png\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/guires.uk\/newsroom\/wp-content\/uploads\/2019\/11\/image-116.png\")
<\/figure><\/div>\n\n\n\n