Arindam Pal

Backend Developer • Java • Spring Boot | Oracle Certified Cloud Architect.
@Infosys Certified Java Programmer & Developer
Creative Head at @cssbitl

[Arindam Pal @a_random_pal]

Day 30 of #summerskillup - @geeksforgeeks

Today's quiz taught:

• Normalization: Scales features to speed up Gradient Descent convergence.
• Gradient Descent: Minimizes MSE using gradients of model parameters.

Apply: https://t.co/dlVY55x3Jc

#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 29 of #SummerSkillUp with @geeksforgeeks

Today's quiz covered:
• Cost Function & R²: Error measure; high R² = better fit.
• Gradient Descent & Learning Rate (α): Minimizes error; α controls learning speed.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 28 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Slope (m): Shows how much y changes when x increases by 1 unit.
• Direction: Positive m means y increases; negative m means y decreases.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 27 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Heteroscedasticity = changing residual variance; R² = variance explained
• LASSO (L1) shrinks coefficients and selects features

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 26 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Ridge Regression helps handle multicollinearity by shrinking coefficients
• Residual analysis checks whether regression assumptions are satisfied

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 25 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Isolation Forest detects outliers in high-dimensional data
• Polynomial & interaction features help capture complex feature-target relationships

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 24 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Decision boundaries separate classes; Decision Trees handle complexity.
• Multi-label classification allows multiple labels; feature extraction finds key features.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 23 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Hyperparameters are defined before training & tuned to improve model performance.
• Grid Search tests all combinations; Bayesian Optz learns from past trials.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 22 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Cross-validation improves model reliability and reduces overfitting.
• K-Fold averages folds; Stratified preserves balance; LOOCV leaves one out.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 21 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• ROC-AUC evaluates binary classifiers across thresholds.
• AUC=0.5 indicates random classification performance.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 20 of #summerskillup -
@geeksforgeeks
Today's quiz taught:
• F1 Score is most useful when there is class imbalance.
• F1 balances precision and recall using the harmonic mean.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 19 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Precision: Fraction of correctly predicted positives among all predicted positives.
• Recall is preferred when false negatives are more harmful.

Apply: https://t.co/dlVY55x3Jc

#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 18 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Alpha controls regularization strength; higher values increase shrinkage
• Ridge (L2) reduces coefficient magnitudes to handle multicollinearity
Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 17 of #summerskillup -
@geeksforgeeks

Today's quiz taught:
• resample() converts time series data across frequencies (daily → monthly)
• Lasso (L1) regularization reduces overfitting and can set feature coefficients to zero

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 16 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• PCA reduces dimensionality while retaining most of data's variance.
• t-SNE help visualize high-dimensional data by revealing clusters and local relationships.

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 15 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Isolation Forest detects outliers in high-dimensional data
• Polynomial & interaction features help capture complex feature-target relationships

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 14 of #summerskillup - @geeksforgeeks

Today's quiz taught:
• Median and IQR are robust against outliers
• Log transforms reduce skewness; violin plots and heatmaps reveal patterns

Apply: https://t.co/dlVY55x3Jc
#skillupwithgfg

[Arindam Pal @a_random_pal]

Day 13 of #skillup by
@geeksforgeeks
Today's quiz taught:
• EDA reveals patterns & data issues
• Bivariate analysis studies two variables; IQR detects outliers
• Imputation handles missing values; log transform reduces skewness#skillupwithgfg #summerskillup #summerskillup2026

[Arindam Pal @a_random_pal]

Day 12 of #skillup by
@geeksforgeeks
Today's quiz taught:

• Filter methods use tests like Chi-Square
• Wrapper methods use Forward and Backward Selection
• Feature Selection reduces dimensionality
#skillupwithgfg #summerskillup #summerskillup2026

[Arindam Pal @a_random_pal]

Day 11 of #skillup by
@geeksforgeeks
Today's quiz taught:

• Feature Selection picks relevant features
• Feature Creation builds new features using patterns
• Min-Max Scaling rescales values to a range of 0–1

#skillupwithgfg #summerskillup #summerskillup2026