Deepshikha Datta
Deepshikha Datta has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Bimal Das
Department of Chemical Engineering National Institute of Technology Durgapur Durgapur India
9 shared publications Gokulan Ravindiran
Department of Civil Engineering, Dayananda Sagar College of Engineering, Bengaluru, 560111, Karnataka, India
3 shared publicationsSayantan Sarkar
Department of Chemical Engineering, National Institute of Technology Durgapur, Durgapur, India
3 shared publicationsDharmana Tejeswararao
Department of Basic Science and Humanities , GMR Institute of Technology , Rajam 532127 , Andhra Pradesh , India
2 shared publicationsDivakar Pamanji
Mother Theresa Institute Of Engineering and Technology
2 shared publications- Karthick Kanagarathinam
Department of Electrical and Electronics Engineering, GMR Institute of Technology, Rajam, India
2 shared publications
Latest Publications
- 3 March 2026
- Published by Springer Nature in Environmental Monitoring and Assessment
- Vol. 198 (3)
A cluster-aware synthetic resampling and machine learning framework for multi-class Air Quality Index classification
The Air Quality Index (AQI) is a widely used indicator for assessing ambient pollution levels and associated health risks. However, AQI classification using data-driven approaches is often challenged by severe class imbalance, where extreme pollution categories such as Poor and Very Poor are sparsely represented. In this study, machine learning models were developed for multi-class AQI classification using a dataset comprising major air pollutants (PM₂.₅, PM₁₀, NO, NO₂, NOx, SO₂, CO, O₃, and volatile organic compounds) along with key meteorological variables, including relative humidity, wind speed, wind direction, solar radiation, and vertical wind speed. To improve model robustness, data preprocessing techniques such as handling missing values, distribution normalisation, and cyclical encoding of wind direction were applied. To address the pronounced class imbalance, a cluster-aware synthetic oversampling (CASO) framework was implemented by integrating random oversampling, Edited Nearest Neighbours (ENN), KMeans-SMOTE, and final class equalisation. Five machine learning models-Logistic Regression, Decision Tree, Random Forest, XGBoost, and LightGBM-were evaluated using accuracy, balanced accuracy, and macro-averaged F1-score. The experimental results demonstrate that ensemble gradient-boosting models, particularly LightGBM and XGBoost, consistently outperform other classifiers, achieving the highest test balanced accuracy values (≥0.96) after resampling. The findings confirm that combining cluster-aware synthetic resampling with ensemble learning significantly enhances AQI classification performance under severe data imbalance, offering a reliable and interpretable framework for air quality assessment.
- 3 March 2026
- Published by Springer Nature in Environmental Monitoring and Assessment
- Vol. 198 (3)
A cluster-aware synthetic resampling and machine learning framework for multi-class Air Quality Index classification
The Air Quality Index (AQI) is a widely used indicator for assessing ambient pollution levels and associated health risks. However, AQI classification using data-driven approaches is often challenged by severe class imbalance, where extreme pollution categories such as Poor and Very Poor are sparsely represented. In this study, machine learning models were developed for multi-class AQI classification using a dataset comprising major air pollutants (PM₂.₅, PM₁₀, NO, NO₂, NOx, SO₂, CO, O₃, and volatile organic compounds) along with key meteorological variables, including relative humidity, wind speed, wind direction, solar radiation, and vertical wind speed. To improve model robustness, data preprocessing techniques such as handling missing values, distribution normalisation, and cyclical encoding of wind direction were applied. To address the pronounced class imbalance, a cluster-aware synthetic oversampling (CASO) framework was implemented by integrating random oversampling, Edited Nearest Neighbours (ENN), KMeans-SMOTE, and final class equalisation. Five machine learning models-Logistic Regression, Decision Tree, Random Forest, XGBoost, and LightGBM-were evaluated using accuracy, balanced accuracy, and macro-averaged F1-score. The experimental results demonstrate that ensemble gradient-boosting models, particularly LightGBM and XGBoost, consistently outperform other classifiers, achieving the highest test balanced accuracy values (≥0.96) after resampling. The findings confirm that combining cluster-aware synthetic resampling with ensemble learning significantly enhances AQI classification performance under severe data imbalance, offering a reliable and interpretable framework for air quality assessment.
- 9 June 2025
- Published by Springer Nature in Asian Journal of Civil Engineering
- Vol. 26 (8) , 3379-3400
An advanced machine learning framework for predicting climate warming from greenhouse gas emissions
- 9 June 2025
- Published by Springer Nature in Asian Journal of Civil Engineering
- Vol. 26 (8) , 3379-3400
An advanced machine learning framework for predicting climate warming from greenhouse gas emissions
- 17 February 2025
- Published by Walter de Gruyter GmbH in International Journal of Chemical Reactor Engineering
- Vol. 23 (6) , 729-738
Synthesis of an excellent red phosphor and photo luminescence studies of Eu 3+ activated double perovskite phosphors NaSrBi 1- x Eu x WO 6 ( x = 0.0–0.24)
The sodium magnesium contains double perovskite structure compound i.e. NaSrBi 1- x Eu x WO 6 is synthesized by using citrate – gel method. This double perovskite structure compound is prepared at different concentrations i.e. ( x = 0.0–0.24) at different temperatures 500, 600, 700 °C for 5 h. Further, the prepared samples are characterized through different tools like XRD (X-ray diffractro meter), UV Visible Diffuse Reflectance Spectra, Photo luminescence spectra, FE SEM and CIE Software and the prepared sample shows different structures in various concentrations by using X-ray diffractrometer. Similarly, lattice parameters, peak positions are explained for all various concentrated samples at 700 °C sintered temperature for 5 h. However, the samples which are sintered at 700 °C for 5 h are excited at different wavelengths i.e. 395, 465, 540 nm, the corresponding emission intensities are also recorded. All the concentrated samples exhibits towards high intensity peaks at 595, 614 nm wavelengths in their emission spectra. The sample x = 0.12 shows highest red emission intensity out of all concentrations in the emission spectra. 5 D 0 - 7 F 1 & 5 D 0 - 7 F 2 are the major transitions formed in the emission spectra for every concentrated sample. In the same way the structural morphology is identified clearly by using FE SSEM. The CIE color coordinators are calculated and identified as very close to NTSC coordinators of red phosphor. Finally, the color purity is also calculated to analyze the quality of red light for the prepared red phosphor which is used in different W-LEDs applications.
- 17 February 2025
- Published by Walter de Gruyter GmbH in International Journal of Chemical Reactor Engineering
- Vol. 23 (6) , 729-738
Synthesis of an excellent red phosphor and photo luminescence studies of Eu 3+ activated double perovskite phosphors NaSrBi 1- x Eu x WO 6 ( x = 0.0–0.24)
The sodium magnesium contains double perovskite structure compound i.e. NaSrBi 1- x Eu x WO 6 is synthesized by using citrate – gel method. This double perovskite structure compound is prepared at different concentrations i.e. ( x = 0.0–0.24) at different temperatures 500, 600, 700 °C for 5 h. Further, the prepared samples are characterized through different tools like XRD (X-ray diffractro meter), UV Visible Diffuse Reflectance Spectra, Photo luminescence spectra, FE SEM and CIE Software and the prepared sample shows different structures in various concentrations by using X-ray diffractrometer. Similarly, lattice parameters, peak positions are explained for all various concentrated samples at 700 °C sintered temperature for 5 h. However, the samples which are sintered at 700 °C for 5 h are excited at different wavelengths i.e. 395, 465, 540 nm, the corresponding emission intensities are also recorded. All the concentrated samples exhibits towards high intensity peaks at 595, 614 nm wavelengths in their emission spectra. The sample x = 0.12 shows highest red emission intensity out of all concentrations in the emission spectra. 5 D 0 - 7 F 1 & 5 D 0 - 7 F 2 are the major transitions formed in the emission spectra for every concentrated sample. In the same way the structural morphology is identified clearly by using FE SSEM. The CIE color coordinators are calculated and identified as very close to NTSC coordinators of red phosphor. Finally, the color purity is also calculated to analyze the quality of red light for the prepared red phosphor which is used in different W-LEDs applications.
- 7 January 2025
- Published by Springer Nature in Journal of the Iranian Chemical Society
- Vol. 22 (2) , 445-456
Europium activated bismuth containing double perovskite catalyzed synthesis of coumarin derivatives and its cytotoxicity evaluation
- 7 January 2025
- Published by Springer Nature in Journal of the Iranian Chemical Society
- Vol. 22 (2) , 445-456
Europium activated bismuth containing double perovskite catalyzed synthesis of coumarin derivatives and its cytotoxicity evaluation
- 27 November 2024
- Published by Springer Nature in Journal of The Institution of Engineers (India): Series B
- Vol. 106 (6) , 1701-1710
Sensing Performance Analysis of Passivated β-(AlxGa1-x)2 O3/Ga2O3 Schottky Diode Gas Sensor with Catalytic Metals at High Temperature
- 27 November 2024
- Published by Springer Nature in Journal of The Institution of Engineers (India): Series B
- Vol. 106 (6) , 1701-1710
Sensing Performance Analysis of Passivated β-(AlxGa1-x)2 O3/Ga2O3 Schottky Diode Gas Sensor with Catalytic Metals at High Temperature
- 11 November 2024
- Published by Taylor & Francis in Indian Chemical Engineer
- Vol. 67 (1) , 49-68
A sustainable approach to enhance biodegradability of recycled LDPE with starch biofillers derived from potato peels and rice husk ash*
- 11 November 2024
- Published by Taylor & Francis in Indian Chemical Engineer
- Vol. 67 (1) , 49-68
A sustainable approach to enhance biodegradability of recycled LDPE with starch biofillers derived from potato peels and rice husk ash*
- 17 October 2024
- Published by Taylor & Francis in Indian Chemical Engineer
- Vol. 67 (1) , 16-30
Drinking water treatment through Optimising Phosphate Coagulation collected from four villages in Srikakulam, India: a Response Surface Methodology approach
- 17 October 2024
- Published by Taylor & Francis in Indian Chemical Engineer
- Vol. 67 (1) , 16-30
Drinking water treatment through Optimising Phosphate Coagulation collected from four villages in Srikakulam, India: a Response Surface Methodology approach
- 27 September 2024
- Published by Springer Nature in Biomass Conversion and Biorefinery
- Vol. 15 (22) , 29049-29066
Fabrication and optimization of working electrodes by using nanosilica extracted from rice straw
- 27 September 2024
- Published by Springer Nature in Biomass Conversion and Biorefinery
- Vol. 15 (22) , 29049-29066
Fabrication and optimization of working electrodes by using nanosilica extracted from rice straw
- 27 April 2024
- Published by Springer Nature in Clean Technologies and Environmental Policy
- Vol. 27 (1) , 97-115
Experimental elucidation of performance of four extracting solvents for reclamation of spent lubricating oil using extraction–flocculation process
- 27 April 2024
- Published by Springer Nature in Clean Technologies and Environmental Policy
- Vol. 27 (1) , 97-115
Experimental elucidation of performance of four extracting solvents for reclamation of spent lubricating oil using extraction–flocculation process
- 1 February 2024
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 102 (6) , 2068-2082
Assessing the efficacy of three bio‐based flocculants in the reclamation of spent lubricating oil
The current study encompasses a comprehensive assessment of three biopolymeric flocculants on the overall performance of recycling waste lubricating oil to achieve a higher percentage recovery, flocculation efficacy, and better quality of recovered base oil. The findings reveal that with experimental conditions such as (i) mixing time of 80 min; (ii) agitation speed of 400 rpm; (iii) reaction temperature 50°C; (iv) solvent to waste oil ratio 3:1 g/g; and (v) flocculant dosage 1 g/ kg of solvents, 1‐butanol and sodium alginate gives highest percentage yield of 91% followed by corn starch of 89.10% and xanthan gum of 87.18% as bio‐polymer flocculant. The effects of various process parameters of bio‐flocculants on flocculation efficiency are expounded. With the process parameters of (i) initial pH of 5.9, 6.0, and 6.2; (ii) mixing time − 59, 60, and 63 min; and (iii) solution temperature of 59, 60, and 62.2°C, maximum flocculation efficacy (% sludge removal) of 16.24%, 13.01%, and 14.09% were attained for the cases of refined oil treated with sodium alginate, corn starch, and xanthan gum, respectively. Results also reveal that the physicochemical properties of refined base oil treated with 1‐butanol and sodium alginate as bio flocculant are almost close to the virgin lubricating oil. The optimum recovery of high‐quality base oil with the adoption of green technology and solvent–bio flocculant combination can mitigate the environmental impact of waste oil and create an energy‐efficient sustainable condition for the regeneration of re‐refined base oil.
- 1 February 2024
- Published by Wiley in The Canadian Journal of Chemical Engineering
- Vol. 102 (6) , 2068-2082
Assessing the efficacy of three bio‐based flocculants in the reclamation of spent lubricating oil
The current study encompasses a comprehensive assessment of three biopolymeric flocculants on the overall performance of recycling waste lubricating oil to achieve a higher percentage recovery, flocculation efficacy, and better quality of recovered base oil. The findings reveal that with experimental conditions such as (i) mixing time of 80 min; (ii) agitation speed of 400 rpm; (iii) reaction temperature 50°C; (iv) solvent to waste oil ratio 3:1 g/g; and (v) flocculant dosage 1 g/ kg of solvents, 1‐butanol and sodium alginate gives highest percentage yield of 91% followed by corn starch of 89.10% and xanthan gum of 87.18% as bio‐polymer flocculant. The effects of various process parameters of bio‐flocculants on flocculation efficiency are expounded. With the process parameters of (i) initial pH of 5.9, 6.0, and 6.2; (ii) mixing time − 59, 60, and 63 min; and (iii) solution temperature of 59, 60, and 62.2°C, maximum flocculation efficacy (% sludge removal) of 16.24%, 13.01%, and 14.09% were attained for the cases of refined oil treated with sodium alginate, corn starch, and xanthan gum, respectively. Results also reveal that the physicochemical properties of refined base oil treated with 1‐butanol and sodium alginate as bio flocculant are almost close to the virgin lubricating oil. The optimum recovery of high‐quality base oil with the adoption of green technology and solvent–bio flocculant combination can mitigate the environmental impact of waste oil and create an energy‐efficient sustainable condition for the regeneration of re‐refined base oil.