Improvement of Jensen, Jensen-Steffensen's, and Jensen's functionals related inequalities for various types of convexity

Basic Information

• Paper ID: 2501.00793
• Title: Improvement of Jensen, Jensen-Steffensen's, and Jensen's functionals related inequalities for various types of convexity
• Author: Shoshana Abramovich
• Classification: math.FA (Functional Analysis)
• Publication Date: January 1, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2501.00793

Abstract

This paper investigates improvements of Jensen inequalities, Jensen-Steffensen inequalities, and Jensen functional-related inequalities for uniformly convex functions, Φ-convex functions, and superquadratic functions.

Research Background and Motivation

Problem Background

Jensen's inequality is a fundamental inequality in convex analysis with widespread applications in mathematical analysis, probability theory, information theory, and other fields. While the classical Jensen inequality applies to convex functions, more precise inequality estimates can be obtained for function classes with special properties, such as uniformly convex functions, Φ-convex functions, and superquadratic functions.

Research Motivation

1. Theoretical Completeness: Existing literature primarily focuses on Jensen-type inequalities for strongly convex functions, lacking systematic research on broader function classes
2. Technical Extension: Generalizing existing technical methods to uniformly convex, Φ-convex, and superquadratic function classes
3. Inequality Improvement: Obtaining more precise estimates than classical Jensen inequalities, particularly improved versions containing additional terms

Limitations of Existing Methods

• Reference 5 only addresses strongly convex functions
• The superquadratic function results in Reference 4 require extension to uniform convexity
• Difference estimates for Jensen functionals need to be established under more general coefficient conditions

Core Contributions

1. Extended Jensen-Steffensen Inequalities: Generalized the results on superquadratic functions from Reference 4 to uniformly convex functions
2. Generalized Jensen-type Results: Extended the results on strongly convex functions from Reference 5 to uniformly convex, Φ-convex, and superquadratic functions
3. Improved Jensen Functional Difference Estimates: Refined the results on differences between two Jensen functionals from Reference 2
4. Established a Unified Technical Framework: Unified treatment of three function classes through inequalities of the form $f(x) - f(u) \geq \varphi(u)(x-u) + \Phi(|x-u|)$

Methodology Details

Core Function Class Definitions

1. Uniformly Convex Functions

Definition 1: A function $f: [a,b] \to \mathbb{R}$ is called Φ-uniformly convex if for all $x,y \in [a,b]$, $t \in [0,1]$: $tf(x) + (1-t)f(y) \geq f(tx + (1-t)y) + t(1-t)\Phi(|x-y|)$

When $\Phi \geq 0$, it is called uniformly convex with modulus $\Phi$.

2. Φ-Convex Functions

Definition 2: A function $f$ is called Φ-convex if for all $x,y \in [a,b]$, $t \in [0,1]$: $tf(x) + (1-t)f(y) + t\Phi((1-t)|x-y|) + (1-t)\Phi(t|x-y|) \geq f(tx + (1-t)y)$

3. Superquadratic Functions

Definition 3: A function $\varphi: [0,\infty) \to \mathbb{R}$ is called superquadratic if for all $x \geq 0$ there exists a constant $C_x$ such that: $\varphi(y) \geq \varphi(x) + C_x(y-x) + \varphi(|y-x|)$

Technical Innovations

1. H-Superadditivity Concept

Definition 4: A function $g: [a,b] \to \mathbb{R}$ is called H-superadditive if for all $x \leq y$: $g(y) - g(x) \geq H(y-x)$

This concept generalizes traditional superadditivity and provides technical tools for handling uniformly convex functions.

2. Unified Inequality Framework

All three function classes satisfy a basic inequality of the form: $f(x) - f(u) \geq \varphi(u)(x-u) + \Phi(|x-u|)$

where:

• Uniformly convex functions: $\varphi = f'$, $\Phi \geq 0$
• Φ-convex functions: $\Phi$ replaced by $(-\Phi)$, $f'$ replaced by $\varphi$
• Superquadratic functions: $\Phi$ replaced by $f$

Main Technical Results

Lemma 2 (Key Lemma for Uniformly Convex Functions)

If $f$ is continuously differentiable on $[a,b]$ and $f'$ is $\Phi'$-superadditive, then the function: $D(y) = f(y) - f(z) - f'(z)(y-z) - \Phi(|y-z|)$ is non-negative on $[a,b]$, non-increasing on $[a,z)$, and non-decreasing on $[z,b)$.

Experimental Setup

Theoretical Verification Methods

This paper employs pure theoretical analysis methods, verifying results through the following steps:

1. Constructive Proofs: Proving inequalities through explicit construction of auxiliary functions
2. Case-by-Case Analysis: Discussing three cases based on the relative positions of parameter $c$ and sequence $\{x_i\}$
3. Jensen-Steffensen Coefficient Verification: Verifying that coefficients satisfy $0 \leq \sum_{i=1}^j a_i \leq \sum_{i=1}^n a_i$

Typical Examples

Example 1: The function $f(x) = x^n$ ($x \geq 0$, $n = 2,3,...$) is uniformly convex with modulus $\Phi(x) = x^n$.

Experimental Results

Main Theoretical Results

Theorem 6 (Jensen-Steffensen-type Inequality for Uniformly Convex Functions)

For a uniformly convex function $f$, under Jensen-Steffensen coefficient conditions: $f(c) + f'(c)(x-c) + \frac{1}{A_n}\sum_{i=1}^n a_i\Phi(|x_i - c|) \leq \frac{1}{A_n}\sum_{i=1}^n a_i f(x_i)$

In particular, when $c = x$: $\frac{1}{A_n}\sum_{i=1}^n a_i f(x_i) \geq f(x) + \frac{1}{A_n}\sum_{i=1}^n a_i\Phi(|x_i - x|)$

Theorems 7-18 (Jensen-type Results for Three Function Classes)

For each function class, four theorems are established, including:

• Improvements of basic Jensen-type inequalities
• Generalized versions with parameters $\lambda_i$
• Reverse forms of Jensen inequalities
• Mixed-type results

Theorems 19-20 (Improvements of Jensen Functional Differences)

Generalizing results from Reference 2 to more general coefficient conditions, where $m^*$ and $M^*$ are defined as: $m^* = \min_{1 \leq i \leq n}\{m_i, \bar{m}_i\}, \quad M^* = \max_{1 \leq i \leq n}\{m_i, \bar{m}_i\}$

Technical Breakthroughs

1. Generalization of Coefficient Conditions: Extended from non-negative coefficients to Jensen-Steffensen coefficients
2. Extension of Function Classes: Extended from strongly convex functions to three more general function classes
3. Refinement of Estimates: Obtained more precise estimates containing additional positive terms

Related Work

Analysis of Main References

1. Reference 4 (Abramovich et al., 2010): Established Jensen-Steffensen inequalities for superquadratic functions; this paper extends them to uniformly convex functions
2. Reference 5 (Bradanović & Lovičević, 2024): Studied Jensen-type inequalities for strongly convex functions; this paper generalizes to three function classes
3. Reference 2 (Abramovich & Dragomir, 2008): Investigated Jensen functional differences; this paper improves coefficient conditions
4. Reference 9 (Zalinescu, 1983): Classical reference on uniformly convex function theory

Theoretical Position of This Paper

This paper occupies an important position in convex analysis and inequality theory by:

• Unifying treatment methods for multiple convexity concepts
• Generalizing classical Jensen and Jensen-Steffensen inequalities
• Providing more precise estimation tools for practical applications

Conclusions and Discussion

Main Conclusions

1. Theoretical Unification: Established a unified framework for handling uniformly convex, Φ-convex, and superquadratic functions
2. Inequality Improvements: Obtained more precise estimates than classical results, containing additional positive terms
3. Generalization of Coefficient Conditions: Extended results from non-negative coefficients to Jensen-Steffensen coefficients
4. Jensen Functional Improvements: Established difference estimates for Jensen functionals under more general conditions

Limitations

1. Scope of Application: Results primarily apply to real-valued functions on intervals, not addressing more general spaces
2. Computational Complexity: Some results require verification of complex coefficient conditions
3. Practical Applications: The practical application value of theoretical results requires further exploration

Future Directions

1. Generalization to More General Spaces: Consider Banach spaces or more general topological spaces
2. Numerical Applications: Explore applications in numerical analysis and optimization
3. Probabilistic Applications: Investigate application value in probability theory and statistics

In-Depth Evaluation

Strengths

1. Theoretical Completeness: Systematically addresses three important function classes and establishes a complete theoretical framework
2. Technical Innovation: Introduces the H-superadditivity concept, providing new technical tools
3. Precision of Results: Obtains more precise estimates containing additional terms
4. Methodological Unification: Handles different function types through a unified inequality form
5. Clear Presentation: Well-structured paper with complete and detailed proofs

Weaknesses

1. Lack of Practical Applications: Paper primarily contains theoretical results with few concrete application examples
2. Computational Complexity: Verification of certain conditions may be complex in practice
3. Limited Originality: Primarily extends and generalizes existing techniques with limited original innovation

Impact

1. Theoretical Contribution: Provides new tools and results for convex analysis and inequality theory
2. Foundation for Future Research: Provides foundation for further research in related fields
3. Educational Value: Can serve as reference material for functional analysis and convex analysis courses

Applicable Scenarios

1. Mathematical Theory Research: Convex analysis, real analysis, functional analysis research
2. Optimization Theory: Convergence analysis of convex optimization algorithms
3. Probability and Statistics: Establishment of expectation inequalities and concentration inequalities
4. Numerical Analysis: Error estimation for numerical methods

References

The paper cites 9 important references covering classical and recent results in convex analysis, inequality theory, and related applications, providing a solid foundation for the paper's theoretical development.

Overall Assessment: This is a high-quality theoretical mathematics paper that systematically generalizes and improves Jensen-type inequalities. While primarily technical generalization work, the results are complete and proofs are rigorous, making important contributions to convex analysis theory. The paper's main value lies in theoretical refinement and unification, providing useful tools for subsequent research.