Isha Gupta

In the present age, materials science research is quickly entering a data-driven phase. Various data-driven methods can be applied to the detection of materials and estimation of their performances with innumerable applications. Machine learning is the most powerful method whose current development and challenges are elaborated in materials science. The entrance of machine learning brings a deep upheaval to human civilization and considerably advances science, which has made significant progress in materials science research. In materials science, the design approaches of Machine learning are categorized and elaborated along with probable perceptions for future development. Machine learning can help solve complex problems that would otherwise be impossible to solve. However, machine learning’s potential benefits during the cost of creation of big data require the algorithmic procedures having vast volumes of data of diverse types and after multiple data sources, ranging from material characteristics to sensing data. Furthermore, recent significant applications of machine learning in metals, batteries, photovoltaic materials, and nanomaterials are discussed.

Mango quality and productivity are negatively impacted by mango leaf diseases. Mango leaf disease is difficult to accurately diagnose with the unaided eye. In order to categorize mango leaf infections, researchers have recently used a number of machine learning and computer-aided approaches. It has been asserted that problems with rising overfitting, computational complexity, feature dimensionality, and a lack of features properties limit the performance of these techniques. To solve these issues, the work provided a novel categorization system for mango leaf diseases. Preparing data, choosing features, learning and classifying, and assessing performance make up the four stages of the proposed system. From the categories of both healthy and diseased imagery, we identified 1,536 images. The open Kaggle database was used to obtain the dataset for this investigation. Accuracy and sensitivity are evaluation metrics that identify the design that performs at the highest level. A few additional performance measures have also been used. The top model has a 96.2 sensitivity and a 97.9 accuracy.

Terbium(III)‐doped yttrium aluminate perovskite (YAP:xTb³⁺) (x = 0.01–0.08 mol) was synthesized using a simple gel‐combustion method. Structural elucidations were performed using X‐ray diffraction (XRD) and Rietveld analysis. Fourier‐transform infrared spectral studies validated the efficient synthesis of designed doped samples. Transmission electron microscopic images showed the agglomerated irregular dimensions of the synthesized nanocrystalline materials. When excited at 251 nm, a strong emissive line attributed to ⁵D₄ → ⁷F₅ electronic transition was observed at 545 nm (green emission). The maximum luminescence was found at the optimized concentration (0.05 mol) of Tb³⁺ ions; this emission was quenched by dipolar–dipolar (d–d) interactions. Chromaticity (x and y) and correlated colour temperature parameters were obtained by analysing the emission profiles. Finally, the colour coordinates of nanophosphors were closer to the National Television Standards Committee green coordinates, which replicates their potency in the design and architecture of R‐G‐B‐based white LEDs.