Flood Prediction Using Machine Learning Project

Flood Prediction Using Machine Learning Thesis Ideas

                       Best solution will be guided for PhD and MS thesis ideas along with thesis topics, customised thesis support will be given …. nowhere where you are struck up with in all areas, we provide the research support. Thesis ideas will be shared based on your requirements. Explore our work and stay inspired….

1. Flood Prediction Using Ensemble Machine Learning Model

Keywords:

Flood Prediction, Ensemble Machine Learning, Rainfall, Support Vector Classifier(SVC), K-Nearest Neighbor(KNN), Decision Tree Classifier(DTC), Binary Logistic Regression, Stacked Generalization

            We concentrate on the comparative analysis of different machine learning methods to predict the flood in India. We estimate the methods like KNN, Support Vector Classifier, Decision Tree classifier, Binary Logistic Regression and stacked generalization. We utilize the rainfall dataset to test and train the model. Our stacked generalization methods performs best accuracy rate.

2. A novel framework for addressing uncertainties in machine learning-based geospatial approaches for flood prediction

Keywords:

Flood susceptibility, Machine learning algorithm, Uncertainty analysis, GIS, Remote sensing            Our paper offer an outline to decrease spatial disagreement between four separate and hybridized ML based FMS’s namely RF, KNN, MLP and hybridized genetic algorithm-Gaussian radial basis function- support vector regression (GA-RBF-SVR). The optimized models were developed by merging the outcome of our four methods. Our optimized model exhibit increased accuracy and enhanced spatial agreement by decreasing the number of classification faults.  

3. An Intelligent Early Flood Forecasting and Prediction Leveraging Machine and Deep Learning Algorithms with Advanced Alert System

Keywords:

flood forecasting and prediction; multilayer perceptron (MLP); time series analysis; deep learning (DL); artificial neural network (ANN); decision tree (DT); recurrent neural network (RNN); exponential smoothing

            The goal of our paper is to construct a prediction method based on the exponential smoothing LSTM (ES-LSTM) and RNN to predict hourly rainfall seasons and categorize the rainfall using a methods ANN and DT. We can work with the Historical daily weather dataset to test the efficacy of our suggested method. Our result displays that ES-LSTM and ANN performs best when compare to other models. 

4. Flood susceptible prediction through the use of geospatial variables and machine learning methods

Keywords:

Flooding, Flood susceptibility mapping, Natural hazards

            Our paper uses six ML methods like DT, RF, MLP-NN, AdaBoost, LR and SVM. We used several metrics to estimate our models performance. We can reclassify the flood conditioning factors to estimate them by utilizing Frequency Ratio method. The flood susceptibility method can be provided by utilizing the RF method specifies that the flood susceptibility is high, moderate or low. 

5. Predicting and analyzing food susceptibility using boosting‑based ensemble machine learning algorithms with SHapley Additive exPlanations

Keywords:

NGBoost, LightGBM, CatBoost, SHapley Additive exPlanations

            Our paper improve FMSs Adana province on the Mediterranean coast of Türkiye by utilizing tree based ML classifiers. We used the methods like NGBoost for the first time in FSM study as well as comparing LightGBM and CatBoost versus other methods like RF, GB, XGBoost and AdaBoost. Our AdaBoost and LightGBM method gives the best accuracy. Hence we used ML-based FSMs with the aid of an XAI method i.e. SHAP.

6. Prediction of flood routing results in the Central Anatolian region of Türkiye with various machine learning models

Keywords:

Flood routing, Flood management, Gradient-boosted machine, Central Anatolian

            We aim to compare the presentation of different ML methods like Bagged Tree, Extreme Gradient Boosting, Gradient Boosting Machine, RF, KNN and SVM to predict the flood routing models. In addition predicting the success of tree based method recognized by comparing optimized and default parameters. Our gradient boosted machine is the successful method to estimate flood routing.

7. Flood susceptibility prediction using tree-based machine learning models in the GBA

Keywords:

Tree-based machine learning, Flood susceptibility, SHAP values

            We establish an outline for estimating flood susceptibility in GBA by utilizing tree-based ML and geographical information system method. We utilize the tree based methods like RF, Gradient Boost Tree, Extreme Gradient Boosting and Categorical boosting to train and test the ML methods and the trained method can be utilized to predict susceptibility. The model Shapely additive explanation values have the robust effect on flood susceptibility.

8. Spatial prediction of flash flood susceptible areas using novel ensemble of bivariate statistics and machine learning techniques for ungauged region

Keywords:

Flash flood susceptibility modelling, Ungauged region, Bivariate statistical model, Multivariate statistical model

            The goal of our study is to include novel ensemble method to bivariate statistical methods like ANN, SVM and KNN model. We have established flash flood and geospatial dataset. A geospatial dataset can be generated with aspect, plan curvature, NDVI, Slope, SPI, TWI, LULC and STI in GIS. Every influence factor based on correlation utilizing WOE in E-open source software that can ensemble with ANN, SVM and KNN. Finally our WOE-AN performs best when compare to other methods.

9. Real-Time Flood Prediction System Using Machine Learning Algorithms

Keywords:

Flood prediction system, K Neighbours classifier, Logistic regression, Random forest classifier

            Our paper utilizes six ML methods like ANN, K Neighbors classifier, LR, Support Vector Classifier, DT and Random Forest Classifier all are executed separately to estimate and contrast the performance when predicting the flood possibility. Our Random Forest method gives the best performance accuracy.

10. MLFP: Machine Learning Approaches for Flood Prediction in Odisha State

Keywords:

Flood, Naïve bayes, Accuracy

            Our paper utilizes the rainfall data that can fed into various machine learning methods. Before we do this process first we clean and pre-process the data. Then the dataset for cleaning is divided into train set and test set. Then we have to compare the accuracy of the method to estimate and evaluate. Then at last we choose the best model by comparing the accuracy.