Subscribe
Home Blog Page 9

Shruti Sharma IAS: History Enthusiast to UPSC Topper

By
Kumar Prafull
-
0
shruti sharma ias

A Tale of Perseverance and Purpose

In the realm of India’s civil services, few stories resonate as profoundly as that of Shruti Sharma IAS. Securing All India Rank 1 in the UPSC Civil Services Examination 2021, her journey from a history student to the esteemed Indian Administrative Service is a testament to determination, discipline, and an unwavering commitment to public service.

Early Life and Educational Background

Born in Delhi in 1996, Shruti hails from Bijnor, Uttar Pradesh. She completed her early education at Cambridge Primary School and later at Sardar Patel Vidyalaya, both in Delhi. Her academic prowess led her to pursue History Honours at the prestigious St. Stephen’s College, Delhi University. Furthering her academic journey, she enrolled in Jawaharlal Nehru University (JNU) for a Master’s in Modern History. However, her passion for civil services prompted her to focus entirely on UPSC preparation, leading her to join the Residential Coaching Academy at Jamia Millia Islamia.

The UPSC Journey: Trials and Triumph

Shruti’s path to success was not without challenges. In her first attempt, due to an issue with the medium of instruction, she had to take the Mains examination in Hindi, missing the interview call by just one mark. Undeterred, she analyzed her shortcomings, refined her strategy, and approached her second attempt with renewed vigor. Her perseverance paid off when she topped the exam in 2021.

Preparation Strategy

  • Optional Subject: Shruti chose History as her optional subject, aligning with her academic background.
  • Study Routine: She emphasized consistency over long hours, focusing on quality learning rather than the quantity of study time.
  • Answer Writing: Recognizing the importance of articulation, she practiced answer writing diligently, focusing on clarity, structure, and time management.
  • Mock Tests: Regular participation in mock tests helped her assess her preparation levels and identify areas for improvement.
  • Current Affairs: She stayed updated with daily news, making concise notes to aid in revision and answer enrichment.

Shruti also highlighted the importance of limiting study materials to avoid confusion and emphasized the need for self-understanding to focus on one’s strengths and weaknesses.

UPSC CSE 2021 Performance

Shruti’s meticulous preparation culminated in an impressive performance:

  • Essay (Paper-I): 132
  • General Studies-I (Paper-II): 119
  • General Studies-II (Paper-III): 128
  • General Studies-III (Paper-IV): 108
  • General Studies-IV (Paper-V): 139
  • Optional-I (History) (Paper-VI): 150
  • Optional-II (History) (Paper-VII): 156
  • Written Total: 932
  • Personality Test: 173
  • Final Total: 1105

Beyond Academics

Shruti’s profile is enriched with diverse experiences:

  • Cultural Enthusiast: She enjoys exploring different cultures, reading books, and watching movies.
  • Language Learner: Her curiosity drives her to learn new languages and understand diverse cultures.
  • Social Contributor: Shruti aims to bring positive changes to society and is eager to learn new things to contribute effectively.

Vision as an IAS Officer

Shruti aspires to:

  • Enhance Education: Implement policies that improve access and quality of education.
  • Empower Women: Advocate for women’s rights and create opportunities for their socio-economic upliftment.
  • Strengthen Governance: Promote transparency, accountability, and efficiency in public administration.

Conclusion

Shruti Sharma’s journey underscores the power of perseverance, strategic planning, and unwavering commitment. Her story serves as an inspiration for countless aspirants, proving that with dedication and resilience, achieving the pinnacle of success in the UPSC examination is attainable.

Ishita Kishore IAS: From Corporate to UPSC Rank 1

By
Kumar Prafull
-
0
ishita kishore ias

The Journey of a Relentless Dreamer

In the highly competitive world of UPSC Civil Services Examination, where over 10 lakh candidates apply each year and only a handful make it to the final list, Ishita Kishore carved her name in gold. In 2022, she secured All India Rank 1, proving yet again that perseverance, strategy, and self-belief can achieve the seemingly impossible.

Her story is not just about topping UPSC. It’s a story of falling, rising, realigning, and conquering — a story that will inspire generations of aspirants.

Early Life and Family Background

Ishita Kishore was born in Begumpet, Hyderabad, and brought up in Patna, Bihar, before moving to Delhi NCR. Her father, Wing Commander Sanjay Kishore, served in the Indian Air Force, instilling in her values of discipline, service, and national pride. Her mother, Jyoti Kishore, is a retired school teacher and has been Ishita’s strongest emotional pillar.

Coming from a defence background, Ishita experienced structured environments and frequent relocations, which made her adaptable and emotionally resilient — qualities that would later become her strengths in UPSC preparation.

Academic Excellence at SRCC

Ishita completed her schooling from Air Force Bal Bharati School, New Delhi, where she actively participated in academics and sports. She then went on to study Economics Honours at Shri Ram College of Commerce (SRCC), Delhi University, one of India’s most prestigious institutions.

While SRCC typically opens doors to finance and consultancy careers, Ishita was quietly preparing for a life in public service.

Corporate Career: Ernst & Young

After graduating in 2017, Ishita joined Ernst & Young (EY) as a Risk Analyst, a high-paying job that many SRCC graduates aspire to. But while she enjoyed the challenge, something inside her remained unfulfilled.

She knew her true calling lay in making a tangible difference in people’s lives — and bureaucracy was the medium through which she could do it.

So in March 2019, she resigned from her corporate job and dedicated herself full-time to preparing for UPSC — a decision that many considered risky, but one she never regretted.

The UPSC Journey: Failures Before the Flight

Contrary to the myth of ‘one-time toppers’, Ishita’s journey wasn’t linear. Her first two attempts ended at the Preliminary stage — a phase where lakhs of candidates are filtered out.

But instead of giving up, Ishita refined her preparation strategy. She focused on building her fundamentals, structured her notes better, improved test-taking techniques, and paid special attention to mental well-being.

By her third attempt, she had learned to manage pressure, reduce silly mistakes, and approach the exam holistically.

Ishita Kishore’s UPSC Preparation Strategy

1. Optional Subject – Political Science & International Relations (PSIR)

Ishita chose PSIR because of its overlap with GS Paper II and Essay Paper. She loved reading about diplomacy, Indian polity, and political theory — which made the subject feel less burdensome.

She used sources like:

  • Subhra Ranjan Notes
  • Andrew Heywood’s Political Theory
  • Pavneet Singh’s IR
  • Previous years’ questions for pattern analysis

2. Daily Study Routine

Her day was planned meticulously:

  • 6:00 AM – 8:00 AM: Newspaper reading (The Hindu + Indian Express editorial)
  • 8:00 AM – 9:00 AM: Revision of previous day’s notes
  • 10:00 AM – 1:00 PM: Static GS topics or optional subject
  • 2:00 PM – 4:00 PM: Current Affairs, note-making
  • 5:00 PM – 7:00 PM: Practice mock answers or prelims test
  • 8:00 PM – 9:00 PM: Light reading, analysis, reflection

She emphasized consistency over intensity, studying around 8–9 focused hours per day.

Mains Answer Writing Practice

She joined multiple online platforms for daily answer writing. Initially, her answers were too lengthy or theoretical. But with peer reviews, toppers’ copies analysis, and expert feedback, she learned the art of balancing content with structure.

Her answer mantra:

“Think like a bureaucrat, write like a citizen.”

She also integrated current affairs into GS answers and supported arguments with constitutional provisions, case laws, or data.

Mental Health and Motivation

Like every aspirant, Ishita faced self-doubt, fatigue, and the pressure of expectations. During her low days, she followed:

  • Mindfulness meditation (10 mins daily)
  • Listening to motivational podcasts (e.g., IAS interviews, Gaur Gopal Das)
  • Talking to family, especially her mother, who kept her emotionally grounded

She also maintained a gratitude journal, where she listed 3 things she was grateful for daily. This helped keep perspective during hard times.

UPSC CSE 2022 Final Result and Marks

Ishita Kishore’s hard work paid off in style:

PaperMarks
Essay137
GS Paper I121
GS Paper II130
GS Paper III88
GS Paper IV112
Optional Paper I (PSIR)147
Optional Paper II (PSIR)166
Written Total901
Personality Test (Interview)193
Final Total1094

She topped the list among 933 selected candidates.

Interview Experience

Her interview revolved around:

  • The role of the private sector in governance
  • Women in administration
  • Global geopolitics and India’s soft power
  • Mental health initiatives by government

She answered with poise, clarity, and balance — qualities that truly impressed the panel.

Sports, Leadership, and Beyond the Books

Did you know that Ishita Kishore is a national-level footballer?

She represented her school in the Subroto Cup in 2012 — showing that sports taught her leadership, teamwork, and endurance, all vital for a career in administration.

She also interned with:

  • CRY (Child Rights and You)
  • Taught underprivileged children at Tihar Jail’s shelter home
  • Represented India in the Indo-China Youth Delegation in 2017

Her Vision as an IAS Officer

Ishita has expressed a deep interest in:

  • Education reforms to reduce dropout rates in rural areas
  • Gender equity in employment and politics
  • Mental health awareness in schools and workplaces
  • Public-private partnerships for better urban development

She believes in a governance model that is inclusive, tech-enabled, and compassionate.

Lessons from Ishita Kishore IAS for Aspirants

  • Failing Prelims twice is not the end — it’s part of the process
  • Discipline beats motivation
  • Your background doesn’t define your future
  • Stay curious, stay humble, stay hungry

Final Words

Ishita Kishore IAS is more than a UPSC topper. She is a role model for aspirants, a reformer-in-making, and a symbol of grace under pressure.

Her story tells us that you don’t need a perfect start — you need a strong will to finish.

“You may have to fight a battle more than once to win it.”
– Margaret Thatcher

And Ishita fought. And won.

Aditya Srivastava: From IIT Kanpur to UPSC AIR 1

By
Kumar Prafull
-
0
aditya srivastava ias

A Journey of Perseverance and Purpose

Aditya Srivastava’s ascent to the pinnacle of the UPSC Civil Services Examination in 2023 is a testament to unwavering dedication, strategic preparation, and a deep-seated desire to serve the nation. His journey from the corridors of IIT Kanpur to the esteemed ranks of the Indian Administrative Service (IAS) offers invaluable insights for aspirants and observers alike.

Early Life and Academic Excellence

Born in 1998 in Lucknow, Uttar Pradesh, Aditya was raised in a middle-class family that valued education and integrity. His father, Ajay Srivastava, served as an Assistant Audit Officer in the Comptroller and Auditor General (CAG) office, while his mother, Abha Srivastava, is a homemaker.

Aditya’s academic journey began at City Montessori School (CMS), Aliganj branch, where he consistently excelled, securing 97.8% in his 10th grade and 97.5% in his 12th grade examinations. His stellar performance earned him a place at the prestigious Indian Institute of Technology (IIT) Kanpur, where he pursued a dual degree program, obtaining both B.Tech and M.Tech degrees in Electrical Engineering. He graduated with an impressive GPA of 9.7 in B.Tech and a perfect 10 in M.Tech.

Corporate Stint and the Call to Serve

Upon graduation, Aditya commenced his professional career at Goldman Sachs in Bengaluru, working as an analyst. Despite the lucrative prospects, he felt a compelling urge to contribute more directly to societal development. After 15 months in the corporate sector, he resigned to dedicate himself fully to preparing for the UPSC Civil Services Examination.

The UPSC Journey: Trials and Triumphs

Aditya’s path to success was marked by perseverance:

  • First Attempt (2021): Did not clear the Preliminary Examination.
  • Second Attempt (2022): Secured AIR 236, leading to selection in the Indian Police Service (IPS).
  • Third Attempt (2023): Achieved AIR 1, topping the Civil Services Examination.

His journey underscores the importance of resilience and learning from setbacks.

Strategic Preparation and Study Approach

Aditya’s preparation strategy was a blend of disciplined study and smart work:

  • Foundational Reading: Emphasized NCERT textbooks to build a strong conceptual base.
  • Selective Resources: Focused on standard reference books, ensuring depth over breadth.
  • Answer Writing Practice: Regularly practiced writing answers to enhance articulation and time management.
  • Mock Tests: Participated in test series to simulate exam conditions and receive feedback.
  • Current Affairs: Maintained a daily routine of reading newspapers and following relevant magazines.

He also joined the Mains Guidance Program at ForumIAS to refine his answer-writing skills.

Optional Subject: Electrical Engineering

Leveraging his academic background, Aditya chose Electrical Engineering as his optional subject. His proficiency in the subject contributed significantly to his overall score.

UPSC 2023 Marksheet

Aditya’s scores in the UPSC 2023 examination were as follows:

  • Essay: 117
  • General Studies I: 104
  • General Studies II: 132
  • General Studies III: 95
  • General Studies IV: 143
  • Optional Paper I (Electrical Engineering): 148
  • Optional Paper II (Electrical Engineering): 160
  • Personality Test (Interview): 200
  • Total: 1099

Personality Test Insights

During the interview, Aditya was assessed on various parameters, including current affairs awareness, ethical reasoning, and situational judgment. His composed demeanor and thoughtful responses earned him a commendable score of 200 in the personality test.

Motivation and Vision

Aditya’s decision to join the civil services was driven by a desire to effect meaningful change at the grassroots level. He viewed the IAS as a platform to address systemic challenges and contribute to nation-building. His journey reflects a shift from personal ambition to public service.

Current Status and Future Aspirations

Following his selection, Aditya is undergoing training at the Lal Bahadur Shastri National Academy of Administration (LBSNAA) in Mussoorie. His future postings will provide opportunities to implement his vision for inclusive and effective governance.

Conclusion

Aditya Srivastava’s journey from a promising engineer to the top rank in the UPSC Civil Services Examination exemplifies the power of determination, strategic planning, and a commitment to public service. His story serves as an inspiration to countless aspirants aiming to make a difference through the civil services

Quantum Overfitting and Regularization: Enhancing Generalization in Quantum Models

By
Kumar Prafull
-
0

Table of Contents

  1. Introduction
  2. What Is Overfitting in Machine Learning?
  3. Manifestation of Overfitting in Quantum Models
  4. Sources of Overfitting in Quantum Machine Learning
  5. Variational Quantum Circuits and Model Complexity
  6. Role of Circuit Depth in Overfitting
  7. The Curse of Expressivity in QML
  8. Quantum Generalization Theory: Early Insights
  9. Indicators of Overfitting in Quantum Workflows
  10. Evaluating Generalization on Quantum Devices
  11. Regularization Techniques for Quantum Models
  12. Parameter Norm Penalties (L2 Regularization)
  13. Early Stopping in Quantum Training
  14. Circuit Pruning and Parameter Dropout
  15. Noise-Injection as a Regularizer
  16. Ensemble Learning in Quantum Circuits
  17. Cross-Validation in Quantum Learning
  18. Hybrid Regularization Strategies
  19. Research on Generalization Bounds in QML
  20. Conclusion

1. Introduction

Overfitting occurs when a model performs well on training data but fails to generalize to unseen examples. In quantum machine learning (QML), overfitting arises due to overparameterized variational circuits, excessive entanglement, or insufficient training samples.

2. What Is Overfitting in Machine Learning?

  • High accuracy on training set
  • Poor performance on test/validation data
  • Model “memorizes” instead of learning patterns

3. Manifestation of Overfitting in Quantum Models

  • Low training loss, high validation loss
  • Quantum classifiers that learn noise in training measurements
  • Circuit configurations that are too expressive

4. Sources of Overfitting in Quantum Machine Learning

  • Too many variational parameters
  • Deep quantum circuits on small datasets
  • Poor encoding strategies
  • Insufficient shot counts (measurement noise)

5. Variational Quantum Circuits and Model Complexity

  • Like deep neural networks, VQCs can be overparameterized
  • More parameters → more capacity to memorize noise

6. Role of Circuit Depth in Overfitting

  • Deeper circuits often provide greater expressivity
  • But increase risk of overfitting, especially on small data

7. The Curse of Expressivity in QML

  • Highly expressive circuits can represent arbitrary functions
  • Without regularization, this leads to poor generalization

8. Quantum Generalization Theory: Early Insights

  • Still a developing field
  • Concepts like VC-dimension, Rademacher complexity being adapted for QML
  • Fidelity-based generalization bounds under study

9. Indicators of Overfitting in Quantum Workflows

  • Loss curves diverging after initial convergence
  • Overly sensitive circuit outputs to small input changes
  • Highly unstable gradients

10. Evaluating Generalization on Quantum Devices

  • Use separate validation set with fixed shot budget
  • Monitor variance across multiple runs

11. Regularization Techniques for Quantum Models

  • L2 weight decay
  • Circuit structure restrictions
  • Parameter sparsity

12. Parameter Norm Penalties (L2 Regularization)

Add to cost function:
\[
\mathcal{L}_{ ext{reg}} = \mathcal{L} + \lambda \sum_i heta_i^2
\]
Where \( heta_i \) are circuit parameters.

13. Early Stopping in Quantum Training

  • Stop optimization when validation loss increases
  • Helps prevent convergence to overfitted minima

14. Circuit Pruning and Parameter Dropout

  • Disable certain gates randomly during training
  • Reduce circuit size post-training by removing low-contribution parameters

15. Noise-Injection as a Regularizer

  • Add controlled noise to circuit outputs or parameters
  • Mimics classical dropout

16. Ensemble Learning in Quantum Circuits

  • Average predictions from multiple small VQCs
  • Reduces variance and improves robustness

17. Cross-Validation in Quantum Learning

  • k-fold or leave-one-out strategies adapted for QML
  • Evaluate circuit generality over multiple splits

18. Hybrid Regularization Strategies

  • Combine classical regularizers with quantum-specific constraints
  • e.g., depth limits + weight decay + early stopping

19. Research on Generalization Bounds in QML

  • Ongoing research in quantum PAC learning
  • Fidelity-based loss bounds, entropic capacity metrics
  • Open question: What governs the learnability of QNNs?

20. Conclusion

Quantum models, like their classical counterparts, are prone to overfitting. Regularization techniques such as early stopping, L2 penalties, circuit pruning, and hybrid strategies can enhance generalization, ensuring quantum systems not only learn but also perform reliably on unseen data.

.

Gradient Descent in Quantum Landscapes: Navigating Optimization in Quantum Machine Learning

By
Kumar Prafull
-
0

Table of Contents

  1. Introduction
  2. Understanding Quantum Loss Landscapes
  3. What Is Gradient Descent?
  4. Role of Gradients in Quantum Circuit Training
  5. Challenges Unique to Quantum Landscapes
  6. Variational Quantum Circuits and Cost Minimization
  7. The Barren Plateau Phenomenon
  8. Gradient Estimation Techniques
  9. Parameter-Shift Rule for Gradient Descent
  10. Finite Difference Gradients
  11. Shot Noise and Gradient Variance
  12. Gradient Descent Algorithm for QML
  13. Adaptive Learning Rates and Quantum Optimization
  14. Momentum and Quantum-Aware Gradient Updates
  15. Batch vs Full Gradient Descent in QML
  16. Robustness of Gradient Descent to Noise
  17. Hybrid Optimization Schemes
  18. Visualizing Quantum Loss Landscapes
  19. Future Directions in Quantum Optimization
  20. Conclusion

1. Introduction

Gradient descent is a core algorithm in optimization, including quantum machine learning. It enables parameterized quantum circuits to learn patterns or minimize physical quantities by iteratively adjusting parameters to reduce a cost function.

2. Understanding Quantum Loss Landscapes

  • The cost function in QML is derived from measurement outcomes (e.g., expectation values).
  • The optimization surface is high-dimensional, potentially rugged or flat in places.

3. What Is Gradient Descent?

An iterative algorithm that updates parameters \( heta \) by moving in the direction of negative gradient of a loss function \( L \):
\[
heta \leftarrow heta – \eta
abla L( heta)
\]

4. Role of Gradients in Quantum Circuit Training

  • Gradients indicate how circuit outputs change with parameters
  • Used in hybrid quantum-classical loops to minimize loss

5. Challenges Unique to Quantum Landscapes

  • Barren plateaus: flat regions where gradients vanish
  • Stochasticity from quantum measurements
  • Hardware noise and gate infidelity

6. Variational Quantum Circuits and Cost Minimization

  • VQCs are quantum analogs of neural networks
  • Cost = expectation value of an observable or cross-entropy

7. The Barren Plateau Phenomenon

  • In deep or wide circuits, gradient magnitudes shrink exponentially
  • Makes training inefficient or infeasible without strategies

8. Gradient Estimation Techniques

  • Parameter-shift rule (exact and analytic)
  • Finite differences (approximate)
  • Adjoint methods (experimental)

9. Parameter-Shift Rule for Gradient Descent

For a gate generated by \( G \) with eigenvalues ±1:
\[
rac{\partial}{\partial heta} \langle O
angle = rac{1}{2} \left[\langle O( heta + rac{\pi}{2})
angle – \langle O( heta – rac{\pi}{2})
angle
ight]
\]

10. Finite Difference Gradients

\[
rac{dL}{d heta} pprox rac{L( heta + \epsilon) – L( heta – \epsilon)}{2\epsilon}
\]
Simple but noise-sensitive and not hardware-friendly.

11. Shot Noise and Gradient Variance

  • Arises from finite measurements
  • Reduces accuracy of gradient estimate
  • Mitigation: increase shot count, use variance reduction techniques

12. Gradient Descent Algorithm for QML

  1. Initialize parameters \( heta \)
  2. Compute loss \( L( heta) \)
  3. Estimate \(
    abla L( heta) \)
  4. Update: \( heta \leftarrow heta – \eta
    abla L \)
  5. Repeat until convergence

13. Adaptive Learning Rates and Quantum Optimization

  • Adam optimizer adapts learning rate per parameter
  • Robust to noisy gradients and sparse signals

14. Momentum and Quantum-Aware Gradient Updates

  • Use exponentially weighted averages of gradients
  • Helps escape shallow minima and oscillations

15. Batch vs Full Gradient Descent in QML

  • Batch: use small set of training inputs
  • Full: evaluate cost over entire dataset (costly)

16. Robustness of Gradient Descent to Noise

  • Gradient noise can slow convergence
  • Use noise-resilient optimizers (e.g., SPSA)

17. Hybrid Optimization Schemes

  • Classical model updates combined with quantum gradients
  • Useful in hybrid networks (CNN → QNN → Dense)

18. Visualizing Quantum Loss Landscapes

  • Plot 2D cross-sections of cost function
  • Visualize gradients and landscape curvature

19. Future Directions in Quantum Optimization

  • Natural gradient methods
  • Quantum-aware second-order optimizers
  • Learning-rate schedules based on fidelity

20. Conclusion

Gradient descent remains the foundation for quantum model optimization, despite challenges like barren plateaus and noise. With the help of analytic gradient techniques and adaptive strategies, it powers many hybrid and fully quantum machine learning models in practice today.

.
1...8910...657Page 9 of 657

Subscribe for exclusive content

We have a curated list of the most noteworthy news from all across the globe. With any subscription plan, you get access to exclusive articles that let you stay ahead of the curve.

Subscribe

Copyright Syskool.