1. Influence of temperature-dependent variables on the compressive strength of flyash-based geopolymer using hybridized machine learning algorithm 

Contributors : Ashwin Rawat, Sanjog Chhetri Sapkota, Prasenjit Saha, and Sourav Das

Highlights : 

1. The main goal of this paper is to look at how the amounts of major oxides (Si/Al, Al/Na, Si/Na, and Na/H2O) affect the compressive strength of geopolymer blocks. Further, the role of thermal effects in geopolymer concrete strength is also studied

2. In a laboratory environment, numerous factors can be adjusted and regulated, such as the choice and concentration of the solution, curing conditions, and mixing parameters.It is crucial to remember that the researcher cannot control the percentage of oxides present in the source material.Therefore, this study plays a crucial role in addressing this uncontrollable variable in geopolymer concrete development.

3. This study further validates by using the machine learning algorithm for the prediction of the characteristics.

4. Advanced  hybridized machine learning techniques  with accuracy over 98% is achieved.

5. The development of GUI-based interface helps  engineers to take well-informed decisions 

Status : 

Planning to send in  Construction and building materials (I.F. 7.1)

2. Prediction of particulate matters of Kathmandu valley using attention-based Deep learning Techniques 

• Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), Gated Recurrent Unit (GRU), and Recurrent Neural Network (RNN) architectures are used to train time series deep learning models. Additionally, I actively participated in composing a manuscript for research documentation.

3. Empirical approach for Prediction of Laboratory Peak Shear Stress Along the Cohesive Soil-Geosynthetic Interface using optimized machine learning model 

• A laboratory peak shear stress data  is used in the data .

• Trained using an optimal boosting ML model using Python.

• Explainable behavior of the optimized model for decision making. 

• Comparative study with standalone and ensemble machine learning model.

5. Development of Efficient machine learning models for the evaluation of wastewater parameters

6. Prediction of mechanical properties of Limestone calcined clay composites (LC3) Using advanced machine learning techiques

7. Hybridized explainable ensemble modeling for the assessment of critical strength parameters of Engineered cementitious composites (ECC) 

8. River discharge forecasting and hydropower  potential prediction using deep learning models 

9. prediction of liquefaction potential using advanced hybridized models

10. structural health assessments of piezo sensors embedded concrete 

11. Corrosion monitoring using advanced hybridized  deep models with reinforcing feedback  criteria